feat(providers): refresh curated model catalogs and retire Imagen 4 (#10537)

* feat(providers): refresh Gemini Flash catalogs and pricing * fix(providers): refresh gemini-web Flash catalog * chore(providers): eliminate Gemini 3.5/3.6 Flash models * feat(providers): refresh Perplexity Web model mappings * feat(providers): refresh PromptQL and Notion catalogs * feat(providers): refresh KIE TinyCMS and Conol catalogs * feat(providers): refresh OpenCode Zen catalog * chore(providers): finish Gemini Flash cleanup * chore(providers): retire Google Imagen 4
2026-08-19 05:32:19 +03:00 · 2026-08-19 00:27:46 +09:00
parent 08860f5cae
commit df90591415
88 changed files with 818 additions and 1212 deletions
--- a/README.md
+++ b/README.md
@@ -557,7 +557,7 @@ the current catalog at **[radar.omniroute.online/planos](https://radar.omniroute
 - **🧠 Memory you control** — off by default, opt-in int8 vector quantization + typed decay, per-request `x-omniroute-no-memory`. → [Memory](docs/frameworks/MEMORY.md)
 - **🛡️ Security** — prompt-injection guard on every LLM route (red-team suite), opt-in credential-masking guardrail (redacts leaked API keys/secrets in both directions), free DuckDuckGo last-resort web search, and an optional OIDC login gate for the dashboard (password login always stays available). → [Guardrails](docs/security/GUARDRAILS.md)
 - **🖼️ New endpoints** — `/v1/ocr` (Mistral OCR) and `/v1/audio/translations` (Whisper-style) round out the media surface. → [API Reference](docs/reference/API_REFERENCE.md)
- **🎨 Image / video / audio generation** — one API for media: xAI Grok Imagine & Novita AI video, ComfyUI, Freepik, Adobe Firefly, Microsoft Designer, Google Imagen, Segmind, EdgeTTS. → [API Reference](docs/reference/API_REFERENCE.md)
+- **🎨 Image / video / audio generation** — one API for media: xAI Grok Imagine & Novita AI video, ComfyUI, Freepik, Adobe Firefly, Microsoft Designer, Segmind, EdgeTTS. → [API Reference](docs/reference/API_REFERENCE.md)
 - **🌍 Deployment & ops** — reverse-proxy `basePath`, browser-language auto-detect, per-key device tracking, root-less MITM trust, zh-TW localization. → [Environment](docs/reference/ENVIRONMENT.md)
 - **🤝 More providers & agents** — Cursor Cloud Agent, Grok Build (xAI) with browser + OAuth login, Ollama first-class card, Claude Opus 5 & Sonnet 5, Kimi official partnership (Code/Web/Moonshot), Zed, Requesty, SenseNova, Yuanbao, Agnes AI… and a refreshed **340-provider catalog**. → [Providers](docs/reference/PROVIDER_REFERENCE.md)
 - **📡 Routing transparency** — every response carries an `X-OmniRoute-Decision` header naming the strategy/provider/latency that served it, a new `cache-optimized` combo strategy + Auto-Combo `cacheAffinity` factor route repeat requests back to the connection holding the cached prefix, and a read-only `/v1/auto-combo/{channel}/candidates` endpoint exposes an `auto/*` channel's live candidate pool. → [Auto-Combo](docs/routing/AUTO-COMBO.md)
--- a/open-sse/config/agyModels.ts
+++ b/open-sse/config/agyModels.ts
@@ -2,8 +2,8 @@
 //
 // These models are pinned from the live `:fetchAvailableModels` endpoint
 // (https://daily-cloudcode-pa.googleapis.com/v1internal:fetchAvailableModels) using a
-// real `agy` consumer-OAuth token. The public catalog exposes the upstream Gemini 3.6
-// and 3.5 Flash ids verbatim; the shared Antigravity executor dispatches them unchanged.
+// real `agy` consumer-OAuth token. The public catalog exposes the upstream Gemini 3.7
+// Flash ids verbatim; the shared Antigravity executor dispatches them unchanged.
 //
 // The `agy` provider reuses the `antigravity` executor/translator (identical backend),
 // but keeps its own catalog so the CLI and IDE model surfaces can evolve independently.
@@ -12,11 +12,11 @@
 // they are not chat-callable.

 export const AGY_PUBLIC_MODELS = Object.freeze([
-  // Gemini 3.6 Flash tiers. The live endpoint selects High by default and advertises
-  // all three ids to both the IDE 2.1.1 and CLI 1.1.x clients.
+  // Gemini 3.7 Flash tiers. The live endpoint selects High by default and advertises
+  // all three ids to both the IDE 2.5.5 and CLI 1.1.x clients.
  {
-    id: "gemini-3.6-flash-high",
-    name: "Gemini 3.6 Flash (High)",
+    id: "gemini-3.7-flash-high",
+    name: "Gemini 3.7 Flash (High)",
    contextLength: 1048576,
    maxOutputTokens: 65536,
    supportsReasoning: true,
@@ -24,8 +24,8 @@ export const AGY_PUBLIC_MODELS = Object.freeze([
    toolCalling: true,
  },
  {
-    id: "gemini-3.6-flash-medium",
-    name: "Gemini 3.6 Flash (Medium)",
+    id: "gemini-3.7-flash-medium",
+    name: "Gemini 3.7 Flash (Medium)",
    contextLength: 1048576,
    maxOutputTokens: 65536,
    supportsReasoning: true,
@@ -33,27 +33,8 @@ export const AGY_PUBLIC_MODELS = Object.freeze([
    toolCalling: true,
  },
  {
-    id: "gemini-3.6-flash-low",
-    name: "Gemini 3.6 Flash (Low)",
-    contextLength: 1048576,
-    maxOutputTokens: 65536,
-    supportsReasoning: true,
-    supportsVision: true,
-    toolCalling: true,
-  },
-  // Claude (Antigravity backend).
-  {
-    id: "claude-opus-4-6-thinking",
-    name: "Claude Opus 4.6 (Thinking)",
-    contextLength: 1048576,
-    maxOutputTokens: 65536,
-    supportsReasoning: true,
-    supportsVision: true,
-    toolCalling: true,
-  },
-  {
-    id: "claude-sonnet-4-6",
-    name: "Claude Sonnet 4.6 (Thinking)",
+    id: "gemini-3.7-flash-low",
+    name: "Gemini 3.7 Flash (Low)",
    contextLength: 1048576,
    maxOutputTokens: 65536,
    supportsReasoning: true,
@@ -79,45 +60,6 @@ export const AGY_PUBLIC_MODELS = Object.freeze([
    supportsVision: true,
    toolCalling: true,
  },
-  {
-    id: "gemini-3-flash-agent",
-    name: "Gemini 3.5 Flash (High)",
-    contextLength: 1048576,
-    maxOutputTokens: 65536,
-    supportsReasoning: true,
-    supportsVision: true,
-    toolCalling: true,
-  },
-  {
-    id: "gemini-3.5-flash-low",
-    name: "Gemini 3.5 Flash (Medium)",
-    contextLength: 1048576,
-    maxOutputTokens: 65536,
-    supportsReasoning: true,
-    supportsVision: true,
-    toolCalling: true,
-  },
-  {
-    id: "gemini-3.5-flash-extra-low",
-    name: "Gemini 3.5 Flash (Low)",
-    contextLength: 1048576,
-    maxOutputTokens: 65536,
-    supportsReasoning: true,
-    supportsVision: true,
-    toolCalling: true,
-  },
-  // Gemini 3.7 Flash: single callable public model (upstream exposes only
-  // gemini-3.7-flash-tiered; suffixed tier ids 404). One entry so it does not
-  // collide under the #3696 public-id uniqueness invariant.
-  {
-    id: "gemini-3.7-flash",
-    name: "Gemini 3.7 Flash",
-    contextLength: 1048576,
-    maxOutputTokens: 65536,
-    supportsReasoning: true,
-    supportsVision: true,
-    toolCalling: true,
-  },
  {
    id: "gemini-3.1-flash-lite",
    name: "Gemini 3.1 Flash Lite",
@@ -125,27 +67,23 @@ export const AGY_PUBLIC_MODELS = Object.freeze([
    maxOutputTokens: 65535,
    toolCalling: true,
  },
-  // Gemini 2.5
+  // Claude (Antigravity backend).
  {
-    id: "gemini-2.5-flash-thinking",
-    name: "Gemini 2.5 Flash Thinking",
+    id: "claude-opus-4-6-thinking",
+    name: "Claude Opus 4.6 (Thinking)",
    contextLength: 1048576,
-    maxOutputTokens: 65535,
+    maxOutputTokens: 65536,
    supportsReasoning: true,
+    supportsVision: true,
    toolCalling: true,
  },
  {
-    id: "gemini-2.5-flash",
-    name: "Gemini 2.5 Flash",
+    id: "claude-sonnet-4-6",
+    name: "Claude Sonnet 4.6 (Thinking)",
    contextLength: 1048576,
-    maxOutputTokens: 65535,
-    toolCalling: true,
-  },
-  {
-    id: "gemini-2.5-flash-lite",
-    name: "Gemini 2.5 Flash Lite",
-    contextLength: 1048576,
-    maxOutputTokens: 65535,
+    maxOutputTokens: 65536,
+    supportsReasoning: true,
+    supportsVision: true,
    toolCalling: true,
  },
  // GPT-OSS
@@ -161,6 +99,21 @@ export const AGY_PUBLIC_MODELS = Object.freeze([

 const AGY_PUBLIC_MODEL_IDS = new Set(AGY_PUBLIC_MODELS.map((model) => model.id));
 const AGY_NON_CHAT_MODEL_IDS = new Set(["tab_flash_lite_preview", "tab_jump_flash_lite_preview"]);
+const AGY_RETIRED_MODEL_IDS = new Set([
+  "gemini-3.6-flash-high",
+  "gemini-3.6-flash-medium",
+  "gemini-3.6-flash-low",
+  "gemini-3-flash-agent",
+  "gemini-3.5-flash-extra-low",
+  "gemini-3.5-flash-low",
+  "gemini-3.5-flash-high",
+  "gemini-3.5-flash-medium",
+  "gemini-3.5-flash-preview",
+  "gemini-2.5-pro",
+  "gemini-2.5-flash-thinking",
+  "gemini-2.5-flash",
+  "gemini-2.5-flash-lite",
+]);

 const AGY_CLIENT_VISIBLE_MODEL_NAMES = Object.freeze(
  AGY_PUBLIC_MODELS.reduce<Record<string, string>>((acc, model) => {
@@ -178,5 +131,5 @@ export function isUserCallableAgyModelId(modelId: string): boolean {
 }

 export function isDiscoverableAgyModelId(modelId: string): boolean {
-  return !!modelId && !AGY_NON_CHAT_MODEL_IDS.has(modelId);
+  return !!modelId && !AGY_NON_CHAT_MODEL_IDS.has(modelId) && !AGY_RETIRED_MODEL_IDS.has(modelId);
 }
--- a/open-sse/config/antigravityModelAliases.ts
+++ b/open-sse/config/antigravityModelAliases.ts
@@ -1,7 +1,7 @@
 export const ANTIGRAVITY_PUBLIC_MODELS = Object.freeze([
-  // Gemini 3.7 Flash tiers listed by the current official Antigravity model catalog
-  // alongside the existing Gemini 3.6 tiers. Keep the upstream model ids unchanged so
-  // discovery and execution address the same models selected by the native client.
+  // Gemini 3.7 Flash tiers listed by the current official Antigravity model catalog.
+  // Keep the upstream model ids unchanged so discovery and execution address the same
+  // models selected by the native client.
  {
    id: "gemini-3.7-flash-high",
    name: "Gemini 3.7 Flash (High)",
@@ -20,51 +20,9 @@ export const ANTIGRAVITY_PUBLIC_MODELS = Object.freeze([
    supportsVision: true,
    toolCalling: true,
  },
-  // Gemini 3.6 Flash tiers retained alongside the newer Gemini 3.7 tiers.
  {
-    id: "gemini-3.6-flash-high",
-    name: "Gemini 3.6 Flash (High)",
-    contextLength: 1048576,
-    maxOutputTokens: 65536,
-    supportsReasoning: true,
-    supportsVision: true,
-    toolCalling: true,
-  },
-  {
-    id: "gemini-3.6-flash-medium",
-    name: "Gemini 3.6 Flash (Medium)",
-    contextLength: 1048576,
-    maxOutputTokens: 65536,
-    supportsReasoning: true,
-    supportsVision: true,
-    toolCalling: true,
-  },
-  {
-    id: "gemini-3.6-flash-low",
-    name: "Gemini 3.6 Flash (Low)",
-    contextLength: 1048576,
-    maxOutputTokens: 65536,
-    supportsReasoning: true,
-    supportsVision: true,
-    toolCalling: true,
-  },
-  // Claude (Antigravity backend). The `agy` provider already ships these from the live
-  // :fetchAvailableModels probe (see agyModels.ts) and discussion #3184 confirmed they
-  // are user-callable through the `antigravity` OAuth provider too — same backend.
-  // `antigravity/claude-opus-4-6-thinking` and `antigravity/claude-sonnet-4-6` both work.
-  // They are upstream IDs, so no alias remapping is required.
-  {
-    id: "claude-opus-4-6-thinking",
-    name: "Claude Opus 4.6 (Thinking)",
-    contextLength: 1048576,
-    maxOutputTokens: 65536,
-    supportsReasoning: true,
-    supportsVision: true,
-    toolCalling: true,
-  },
-  {
-    id: "claude-sonnet-4-6",
-    name: "Claude Sonnet 4.6 (Thinking)",
+    id: "gemini-3.7-flash-low",
+    name: "Gemini 3.7 Flash (Low)",
    contextLength: 1048576,
    maxOutputTokens: 65536,
    supportsReasoning: true,
@@ -92,50 +50,6 @@ export const ANTIGRAVITY_PUBLIC_MODELS = Object.freeze([
    supportsVision: true,
    toolCalling: true,
  },
-  // Gemini 3.5 Flash tiers exposed by Antigravity's model selector. Public ids match
-  // fetchAvailableModels and are forwarded upstream unchanged:
-  //   High   -> gemini-3-flash-agent       (displayName: Gemini 3.5 Flash (High))
-  //   Medium -> gemini-3.5-flash-low       (displayName: Gemini 3.5 Flash (Medium))
-  //   Low    -> gemini-3.5-flash-extra-low (displayName: Gemini 3.5 Flash (Low))
-  {
-    id: "gemini-3-flash-agent",
-    name: "Gemini 3.5 Flash (High)",
-    contextLength: 1048576,
-    maxOutputTokens: 65536,
-    supportsReasoning: true,
-    supportsVision: true,
-    toolCalling: true,
-  },
-  {
-    id: "gemini-3.5-flash-low",
-    name: "Gemini 3.5 Flash (Medium)",
-    contextLength: 1048576,
-    maxOutputTokens: 65536,
-    supportsReasoning: true,
-    supportsVision: true,
-    toolCalling: true,
-  },
-  {
-    id: "gemini-3.5-flash-extra-low",
-    name: "Gemini 3.5 Flash (Low)",
-    contextLength: 1048576,
-    maxOutputTokens: 65536,
-    supportsReasoning: true,
-    supportsVision: true,
-    toolCalling: true,
-  },
-  // Gemini 3.7 Flash: Antigravity's live catalog exposes a single upstream id
-  // gemini-3.7-flash-tiered; the suffixed tier ids 404 upstream. Kept as one
-  // callable public model so it does not collide with the #3696 uniqueness invariant.
-  {
-    id: "gemini-3.7-flash",
-    name: "Gemini 3.7 Flash",
-    contextLength: 1048576,
-    maxOutputTokens: 65536,
-    supportsReasoning: true,
-    supportsVision: true,
-    toolCalling: true,
-  },
  {
    id: "gemini-3.1-flash-lite",
    name: "Gemini 3.1 Flash Lite",
@@ -143,25 +57,27 @@ export const ANTIGRAVITY_PUBLIC_MODELS = Object.freeze([
    maxOutputTokens: 65535,
    toolCalling: true,
  },
+  // Claude (Antigravity backend). The `agy` provider already ships these from the live
+  // :fetchAvailableModels probe (see agyModels.ts) and discussion #3184 confirmed they
+  // are user-callable through the `antigravity` OAuth provider too — same backend.
+  // `antigravity/claude-opus-4-6-thinking` and `antigravity/claude-sonnet-4-6` both work.
+  // They are upstream IDs, so no alias remapping is required.
  {
-    id: "gemini-2.5-flash-thinking",
-    name: "Gemini 2.5 Flash Thinking",
+    id: "claude-opus-4-6-thinking",
+    name: "Claude Opus 4.6 (Thinking)",
    contextLength: 1048576,
-    maxOutputTokens: 65535,
+    maxOutputTokens: 65536,
+    supportsReasoning: true,
+    supportsVision: true,
    toolCalling: true,
  },
  {
-    id: "gemini-2.5-flash",
-    name: "Gemini 2.5 Flash",
+    id: "claude-sonnet-4-6",
+    name: "Claude Sonnet 4.6 (Thinking)",
    contextLength: 1048576,
-    maxOutputTokens: 65535,
-    toolCalling: true,
-  },
-  {
-    id: "gemini-2.5-flash-lite",
-    name: "Gemini 2.5 Flash Lite",
-    contextLength: 1048576,
-    maxOutputTokens: 65535,
+    maxOutputTokens: 65536,
+    supportsReasoning: true,
+    supportsVision: true,
    toolCalling: true,
  },
  {
@@ -175,12 +91,6 @@ export const ANTIGRAVITY_PUBLIC_MODELS = Object.freeze([
 ]);

 export const ANTIGRAVITY_MODEL_ALIASES = Object.freeze({
-  // Gemini 3.7 Flash: the live catalog (fetchAvailableModels on daily-cloudcode-pa)
-  // exposes a single upstream id `gemini-3.7-flash-tiered`; the agy CLI maps all
-  // display tiers (high/medium/low) to it. Verified 200 OK with thinking_level and
-  // thinkingBudget configs. The suffixed ids 404 upstream ("Requested entity was not found").
-  // Exposed as ONE callable model (see #3696: public ids must be unique upstream ids).
-  "gemini-3.7-flash": "gemini-3.7-flash-tiered",
  // gemini-3.1-pro-low is not aliased: the upstream accepts it verbatim.
  // gemini-3.1-pro-high: the discovery slot returns HTTP 400 on v1internal;
  // the live upstream id is gemini-pro-agent (see ANTIGRAVITY_PUBLIC_MODELS).
@@ -249,10 +159,19 @@ const ANTIGRAVITY_NON_CHAT_MODEL_IDS = new Set([
 const ANTIGRAVITY_RETIRED_MODEL_IDS = new Set([
  "gemini-3-pro-preview",
  "gemini-3.1-pro",
+  "gemini-3.6-flash-high",
+  "gemini-3.6-flash-medium",
+  "gemini-3.6-flash-low",
+  "gemini-3-flash-agent",
+  "gemini-3.5-flash-extra-low",
+  "gemini-3.5-flash-low",
  "gemini-3.5-flash-high",
  "gemini-3.5-flash-medium",
  "gemini-3.5-flash-preview",
  "gemini-2.5-pro",
+  "gemini-2.5-flash-thinking",
+  "gemini-2.5-flash",
+  "gemini-2.5-flash-lite",
  "gemini-2.5-computer-use-preview-10-2025",
 ]);

@@ -281,7 +200,12 @@ const ANTIGRAVITY_DROPPED_QUOTA_BUCKETS = new Set<string>([
 */
 export function toClientAntigravityQuotaModelId(modelId: string): string | null {
  if (!modelId) return null;
-  if (ANTIGRAVITY_DROPPED_QUOTA_BUCKETS.has(modelId)) return null;
+  if (
+    ANTIGRAVITY_DROPPED_QUOTA_BUCKETS.has(modelId) ||
+    ANTIGRAVITY_RETIRED_MODEL_IDS.has(modelId)
+  ) {
+    return null;
+  }
  return toClientAntigravityModelId(modelId);
 }

--- a/open-sse/config/freeModelCatalog.data.ts
+++ b/open-sse/config/freeModelCatalog.data.ts
@@ -16,27 +16,20 @@ import type { FreeModelBudget } from "./freeModelCatalog.ts";
 * rewrites file timestamps on every deploy, which would report a months-old
 * catalog as "updated today". Bump this whenever the entries below change.
 */
-export const FREE_CATALOG_CURATED_AT = "2026-07-22";
+export const FREE_CATALOG_CURATED_AT = "2026-08-16";

 export const FREE_MODEL_BUDGETS: FreeModelBudget[] = [
  { provider: "agentrouter", modelId: "claude-opus-4-8", displayName: "Claude Opus 4.8", monthlyTokens: 0, creditTokens: 200000000, freeType: "one-time-initial", poolKey: "agentrouter", tos: "caution" },
  { provider: "agentrouter", modelId: "claude-opus-5", displayName: "Claude Opus 5", monthlyTokens: 0, creditTokens: 200000000, freeType: "one-time-initial", poolKey: "agentrouter", tos: "caution" },
  { provider: "agentrouter", modelId: "gpt-5.6-sol", displayName: "GPT-5.6 Sol", monthlyTokens: 0, creditTokens: 200000000, freeType: "one-time-initial", poolKey: "agentrouter", tos: "caution" },
+  { provider: "agy", modelId: "gemini-3.7-flash-high", displayName: "Gemini 3.7 Flash (High)", monthlyTokens: 0, creditTokens: 0, freeType: "keyless", poolKey: "agy", tos: "avoid" },
+  { provider: "agy", modelId: "gemini-3.7-flash-medium", displayName: "Gemini 3.7 Flash (Medium)", monthlyTokens: 0, creditTokens: 0, freeType: "keyless", poolKey: "agy", tos: "avoid" },
+  { provider: "agy", modelId: "gemini-3.7-flash-low", displayName: "Gemini 3.7 Flash (Low)", monthlyTokens: 0, creditTokens: 0, freeType: "keyless", poolKey: "agy", tos: "avoid" },
+  { provider: "agy", modelId: "gemini-pro-agent", displayName: "Gemini 3.1 Pro (High)", monthlyTokens: 0, creditTokens: 0, freeType: "keyless", poolKey: "agy", tos: "avoid" },
+  { provider: "agy", modelId: "gemini-3.1-pro-low", displayName: "Gemini 3.1 Pro (Low)", monthlyTokens: 0, creditTokens: 0, freeType: "keyless", poolKey: "agy", tos: "avoid" },
+  { provider: "agy", modelId: "gemini-3.1-flash-lite", displayName: "Gemini 3.1 Flash Lite", monthlyTokens: 0, creditTokens: 0, freeType: "keyless", poolKey: "agy", tos: "avoid" },
  { provider: "agy", modelId: "claude-opus-4-6-thinking", displayName: "Claude Opus 4.6 (Thinking)", monthlyTokens: 0, creditTokens: 0, freeType: "keyless", poolKey: "agy", tos: "avoid" },
  { provider: "agy", modelId: "claude-sonnet-4-6", displayName: "Claude Sonnet 4.6 (Thinking)", monthlyTokens: 0, creditTokens: 0, freeType: "keyless", poolKey: "agy", tos: "avoid" },
-  { provider: "agy", modelId: "gemini-3.1-pro-low", displayName: "Gemini 3.1 Pro (Low)", monthlyTokens: 0, creditTokens: 0, freeType: "keyless", poolKey: "agy", tos: "avoid" },
-  { provider: "agy", modelId: "gemini-pro-agent", displayName: "Gemini 3.1 Pro (High)", monthlyTokens: 0, creditTokens: 0, freeType: "keyless", poolKey: "agy", tos: "avoid" },
-  { provider: "agy", modelId: "gemini-3.7-flash", displayName: "Gemini 3.7 Flash", monthlyTokens: 0, creditTokens: 0, freeType: "keyless", poolKey: "agy", tos: "avoid" },
-  { provider: "agy", modelId: "gemini-3.6-flash-high", displayName: "Gemini 3.6 Flash (High)", monthlyTokens: 0, creditTokens: 0, freeType: "keyless", poolKey: "agy", tos: "avoid" },
-  { provider: "agy", modelId: "gemini-3.6-flash-medium", displayName: "Gemini 3.6 Flash (Medium)", monthlyTokens: 0, creditTokens: 0, freeType: "keyless", poolKey: "agy", tos: "avoid" },
-  { provider: "agy", modelId: "gemini-3.6-flash-low", displayName: "Gemini 3.6 Flash (Low)", monthlyTokens: 0, creditTokens: 0, freeType: "keyless", poolKey: "agy", tos: "avoid" },
-  { provider: "agy", modelId: "gemini-3-flash-agent", displayName: "Gemini 3.5 Flash (High)", monthlyTokens: 0, creditTokens: 0, freeType: "keyless", poolKey: "agy", tos: "avoid" },
-  { provider: "agy", modelId: "gemini-3.5-flash-low", displayName: "Gemini 3.5 Flash (Medium)", monthlyTokens: 0, creditTokens: 0, freeType: "keyless", poolKey: "agy", tos: "avoid" },
-  { provider: "agy", modelId: "gemini-3.5-flash-extra-low", displayName: "Gemini 3.5 Flash (Low)", monthlyTokens: 0, creditTokens: 0, freeType: "keyless", poolKey: "agy", tos: "avoid" },
-  { provider: "agy", modelId: "gemini-3.1-flash-lite", displayName: "Gemini 3.1 Flash Lite", monthlyTokens: 0, creditTokens: 0, freeType: "keyless", poolKey: "agy", tos: "avoid" },
-  { provider: "agy", modelId: "gemini-2.5-flash", displayName: "Gemini 2.5 Flash", monthlyTokens: 0, creditTokens: 0, freeType: "keyless", poolKey: "agy", tos: "avoid" },
-  { provider: "agy", modelId: "gemini-2.5-flash-thinking", displayName: "Gemini 2.5 Flash Thinking", monthlyTokens: 0, creditTokens: 0, freeType: "keyless", poolKey: "agy", tos: "avoid" },
-  { provider: "agy", modelId: "gemini-2.5-flash-lite", displayName: "Gemini 2.5 Flash Lite", monthlyTokens: 0, creditTokens: 0, freeType: "keyless", poolKey: "agy", tos: "avoid" },
  { provider: "agy", modelId: "gpt-oss-120b-medium", displayName: "GPT-OSS 120B (Medium)", monthlyTokens: 0, creditTokens: 0, freeType: "keyless", poolKey: "agy", tos: "avoid" },
  { provider: "ai21", modelId: "jamba-large-1.7", displayName: "jamba-large-1.7", monthlyTokens: 0, creditTokens: 10000000, freeType: "one-time-initial", poolKey: "ai21", tos: "avoid" },
  { provider: "ai21", modelId: "jamba-mini-2", displayName: "jamba-mini-2", monthlyTokens: 0, creditTokens: 10000000, freeType: "one-time-initial", poolKey: "ai21", tos: "avoid" },
@@ -188,7 +181,6 @@ export const FREE_MODEL_BUDGETS: FreeModelBudget[] = [
  { provider: "gemini", modelId: "gemini-2.5-flash-lite", displayName: "Gemini 2.5 Flash-Lite", monthlyTokens: 0, creditTokens: 0, freeType: "recurring-daily", poolKey: "gemini-free", tos: "caution" },
  { provider: "gemini", modelId: "gemini-3-flash-preview", displayName: "Gemini 3 Flash Preview", monthlyTokens: 0, creditTokens: 0, freeType: "recurring-daily", poolKey: "gemini-free", tos: "caution" },
  { provider: "gemini", modelId: "gemini-3.1-flash-lite", displayName: "Gemini 3.1 Flash-Lite", monthlyTokens: 0, creditTokens: 0, freeType: "recurring-daily", poolKey: "gemini-free", tos: "caution" },
-  { provider: "gemini", modelId: "gemini-3.5-flash", displayName: "Gemini 3.5 Flash", monthlyTokens: 0, creditTokens: 0, freeType: "recurring-daily", poolKey: "gemini-free", tos: "caution" },
  { provider: "glm-cn", modelId: "glm-4-flash", displayName: "GLM-4-Flash", monthlyTokens: 0, creditTokens: 0, freeType: "recurring-uncapped", poolKey: "zhipu-flash-free", tos: "ok" },
  { provider: "glm-cn", modelId: "glm-4.5-flash", displayName: "GLM-4.5-Flash", monthlyTokens: 0, creditTokens: 0, freeType: "recurring-uncapped", poolKey: "zhipu-flash-free", tos: "ok" },
  { provider: "glm-cn", modelId: "glm-4.7-flash", displayName: "GLM-4.7-Flash", monthlyTokens: 0, creditTokens: 0, freeType: "recurring-uncapped", poolKey: "zhipu-flash-free", tos: "ok" },
@@ -369,7 +361,7 @@ export const FREE_MODEL_BUDGETS: FreeModelBudget[] = [
  { provider: "qoder", modelId: "deepseek-v4-pro", displayName: "DeepSeek-V4-Pro", monthlyTokens: 0, creditTokens: 1000000, freeType: "one-time-initial", poolKey: "qoder", tos: "caution" },
  { provider: "qoder", modelId: "deepseek-v4-flash", displayName: "DeepSeek-V4-Flash", monthlyTokens: 0, creditTokens: 1000000, freeType: "one-time-initial", poolKey: "qoder", tos: "caution" },
  { provider: "qoder", modelId: "minimax-m3", displayName: "MiniMax-M3", monthlyTokens: 0, creditTokens: 1000000, freeType: "one-time-initial", poolKey: "qoder", tos: "caution" },
-  { provider: "qwen-web", modelId: "qwen3.8-max", displayName: "Qwen3.8 Max", monthlyTokens: 0, creditTokens: 0, freeType: "keyless", poolKey: "qwen-web", tos: "avoid" },
+  { provider: "qwen-web", modelId: "qwen3.8-max-preview", displayName: "Qwen3.8 Max Preview", monthlyTokens: 0, creditTokens: 0, freeType: "keyless", poolKey: "qwen-web", tos: "avoid" },
  { provider: "qwen-web", modelId: "qwen3.7-max", displayName: "Qwen3.7 Max", monthlyTokens: 0, creditTokens: 0, freeType: "keyless", poolKey: "qwen-web", tos: "avoid" },
  { provider: "qwen-web", modelId: "qwen3.7-plus", displayName: "Qwen3.7 Plus", monthlyTokens: 0, creditTokens: 0, freeType: "keyless", poolKey: "qwen-web", tos: "avoid" },
  { provider: "qwen-web", modelId: "qwen3.6-plus", displayName: "Qwen3.6 Plus", monthlyTokens: 0, creditTokens: 0, freeType: "keyless", poolKey: "qwen-web", tos: "avoid" },
--- a/open-sse/config/geminiRateLimits.json
+++ b/open-sse/config/geminiRateLimits.json
@@ -5,9 +5,6 @@
  "gemini-2-flash-lite": { "rpm": 0, "rpd": 0, "tpm": 0 },
  "gemini-2.5-flash-tts": { "rpm": 3, "rpd": 10, "tpm": 10000 },
  "gemini-2.5-pro-tts": { "rpm": 0, "rpd": 0, "tpm": 0 },
-  "imagen-4-generate": { "rpm": -1, "rpd": 25, "tpm": -1 },
-  "imagen-4-ultra-generate": { "rpm": -1, "rpd": 25, "tpm": -1 },
-  "imagen-4-fast-generate": { "rpm": -1, "rpd": 25, "tpm": -1 },
  "gemma-4-26b-it": { "rpm": 16000, "rpd": 14400, "tpm": 16000 },
  "gemma-4-31b-it": { "rpm": 16000, "rpd": 14400, "tpm": 16000 },
  "gemini-embedding-exp-03-07": { "rpm": 100, "rpd": 1000, "tpm": 30000 },
--- a/open-sse/config/imageRegistry.ts
+++ b/open-sse/config/imageRegistry.ts
@@ -10,7 +10,6 @@ import { SEGMIND_IMAGE_PROVIDER } from "./providers/registry/segmind/imageModels
 import { KIE_IMAGE_MODELS } from "./providers/registry/kie/imageModels.ts";
 import { FREEPIK_IMAGE_PROVIDER } from "./providers/registry/freepik/index.ts";
 import { STABILITY_AI_IMAGE_MODELS } from "./providers/registry/stability-ai/imageModels.ts";
-import { GEMINI_IMAGEN_PROVIDER } from "./providers/registry/gemini/imageModels.ts";
 import { CHEAPERINFERENCE_IMAGE_PROVIDER } from "./providers/registry/cheaperinference/imageModels.ts";
 import {
  ADOBE_FIREFLY_IMAGE_ROUTING_ALIASES,
@@ -378,10 +377,6 @@ export const IMAGE_PROVIDERS: Record<string, ImageProviderConfig> = {
    supportedSizes: ["1024x1024"],
  },

-  // Google AI Studio Imagen family — dedicated :predict endpoint, not generateContent.
-  // See providers/registry/gemini/imageModels.ts for the full rationale.
-  gemini: GEMINI_IMAGEN_PROVIDER,
-
  //Curruntly no models serving
  nebius: {
    id: "nebius",
--- a/open-sse/config/providers/registry/cheaperinference/index.ts
+++ b/open-sse/config/providers/registry/cheaperinference/index.ts
@@ -107,7 +107,12 @@ export const CHEAPERINFERENCE_MODELS: RegistryModel[] = [
    supportsReasoning: true,
    toolCalling: true,
  },
-  { id: "deepseek-v4-flash", name: "DeepSeek V4 Flash", supportsReasoning: true, toolCalling: true },
+  {
+    id: "deepseek-v4-flash",
+    name: "DeepSeek V4 Flash",
+    supportsReasoning: true,
+    toolCalling: true,
+  },
  { id: "deepseek-v4-pro", name: "DeepSeek V4 Pro", supportsReasoning: true, toolCalling: true },
  {
    id: "gemini-2.5-flash",
@@ -117,8 +122,8 @@ export const CHEAPERINFERENCE_MODELS: RegistryModel[] = [
    toolCalling: true,
  },
  {
-    id: "gemini-3-5-flash",
-    name: "Gemini 3.5 Flash",
+    id: "gemini-3.7-flash",
+    name: "Gemini 3.7 Flash",
    supportsVision: true,
    supportsReasoning: true,
    toolCalling: true,
--- a/open-sse/config/providers/registry/cursor/index.ts
+++ b/open-sse/config/providers/registry/cursor/index.ts
@@ -140,22 +140,19 @@ export const cursorProvider: RegistryEntry = {
    { id: "claude-4.6-sonnet-medium", name: "Claude 4.6 Sonnet Medium" },
    { id: "claude-4.6-sonnet-medium-thinking", name: "Claude 4.6 Sonnet Medium Thinking" },
    //
-    { id: "claude-4.5-sonnet", name: "Claude 4.5 Sonnet" },
-    { id: "claude-4.5-sonnet-thinking", name: "Claude 4.5 Sonnet Thinking" },
-    //
    { id: "gemini-3.1-pro", name: "Gemini 3.1 Pro" },
    //
+    { id: "gemini-3.7-flash", name: "Gemini 3.7 Flash" },
    { id: "gemini-3-flash", name: "Gemini 3 Flash" },
    //
-    { id: "grok-4.3", name: "Grok 4.3" },
+    { id: "grok-4.6-medium", name: "Grok 4.6 Medium" },
+    { id: "grok-4.6-fast-medium", name: "Grok 4.6 Fast Medium" },
+    { id: "grok-4.6-high", name: "Grok 4.6 High" },
+    { id: "grok-4.6-fast-high", name: "Grok 4.6 Fast High" },
+    { id: "grok-4.6-xhigh", name: "Grok 4.6 XHigh" },
+    { id: "grok-4.6-fast-xhigh", name: "Grok 4.6 Fast XHigh" },
    //
-    { id: "grok-4.5-medium", name: "Grok 4.5 Medium" },
-    { id: "grok-4.5-fast-medium", name: "Grok 4.5 Fast Medium" },
-    { id: "grok-4.5-high", name: "Grok 4.5 High" },
-    { id: "grok-4.5-fast-high", name: "Grok 4.5 Fast High" },
-    { id: "grok-4.5-xhigh", name: "Grok 4.5 XHigh" },
-    { id: "grok-4.5-fast-xhigh", name: "Grok 4.5 Fast XHigh" },
-    //
-    { id: "kimi-k2.5", name: "Kimi K2.5" },
+    { id: "kimi-k3", name: "Kimi K3" },
+    { id: "kimi-k2.7-code", name: "Kimi K2.7 Code" },
  ],
 };
--- a/open-sse/config/providers/registry/devin/catalog.ts
+++ b/open-sse/config/providers/registry/devin/catalog.ts
@@ -84,14 +84,10 @@ export const DEVIN_MODEL_CATALOG: RegistryModel[] = [
  // Gemini
  { id: "gemini-3-1-pro-high", name: "Gemini 3.1 Pro High", contextLength: 1048576 },
  { id: "gemini-3-1-pro-low", name: "Gemini 3.1 Pro Low", contextLength: 1048576 },
-  { id: "gemini-3-6-flash-high", name: "Gemini 3.6 Flash High" },
-  { id: "gemini-3-6-flash-medium", name: "Gemini 3.6 Flash Medium" },
-  { id: "gemini-3-6-flash-low", name: "Gemini 3.6 Flash Low" },
-  { id: "gemini-3-6-flash-minimal", name: "Gemini 3.6 Flash Minimal" },
-  { id: "gemini-3-5-flash-high", name: "Gemini 3.5 Flash High", contextLength: 1048576 },
-  { id: "gemini-3-5-flash-medium", name: "Gemini 3.5 Flash Medium", contextLength: 1048576 },
-  { id: "gemini-3-5-flash-low", name: "Gemini 3.5 Flash Low", contextLength: 1048576 },
-  { id: "gemini-3-5-flash-minimal", name: "Gemini 3.5 Flash Minimal", contextLength: 1048576 },
+  { id: "gemini-3-7-flash-high", name: "Gemini 3.7 Flash High" },
+  { id: "gemini-3-7-flash-medium", name: "Gemini 3.7 Flash Medium" },
+  { id: "gemini-3-7-flash-low", name: "Gemini 3.7 Flash Low" },
+  { id: "gemini-3-7-flash-minimal", name: "Gemini 3.7 Flash Minimal" },
  // Grok
  { id: "grok-4-5-high", name: "Grok 4.5 High", contextLength: 500000 },
  { id: "grok-4-5-medium", name: "Grok 4.5 Medium", contextLength: 500000 },
--- a/open-sse/config/providers/registry/gemini/imageModels.ts
+++ b/open-sse/config/providers/registry/gemini/imageModels.ts
@@ -1,32 +0,0 @@
-/**
- * Google AI Studio (Gemini API) Imagen family image-generation provider entry.
- *
- * Uses the dedicated `:predict` endpoint (handled by format "google-imagen"), NOT
- * generateContent — so only imagen-* models belong here; gemini flash-image /
- * nano-banana route through /v1/chat/completions instead. The models are also
- * surfaced live via ListModels; this seed makes them addressable on
- * /v1/images/generations. Note: Imagen requires a billing-enabled Google project —
- * free-tier keys get 403 / quota 0. The handler builds `{baseUrl}/{model}:predict`.
- *
- * Extracted out of imageRegistry.ts (which sits right at the 800-line file-size
- * cap) so the catalog lives in its own semantic family module, following the same
- * pattern as `providers/registry/stability-ai/imageModels.ts` and
- * `providers/registry/segmind/imageModels.ts`. Co-located with the existing
- * `gemini/index.ts` chat-provider entry — same provider id, different
- * modality/consumer (chat registry vs image registry), mirroring the
- * `kie/index.ts` + `kie/imageModels.ts` split.
- */
-export const GEMINI_IMAGEN_PROVIDER = {
-  id: "gemini",
-  alias: "gemini",
-  baseUrl: "https://generativelanguage.googleapis.com/v1beta/models",
-  authType: "apikey",
-  authHeader: "x-goog-api-key",
-  format: "google-imagen",
-  models: [
-    { id: "imagen-4.0-generate-001", name: "Imagen 4" },
-    { id: "imagen-4.0-ultra-generate-001", name: "Imagen 4 Ultra" },
-    { id: "imagen-4.0-fast-generate-001", name: "Imagen 4 Fast" },
-  ],
-  supportedSizes: ["1024x1024", "1792x1024", "1024x1792"],
-};
--- a/open-sse/config/providers/registry/gemini/index.ts
+++ b/open-sse/config/providers/registry/gemini/index.ts
@@ -22,14 +22,14 @@ export const geminiProvider: RegistryEntry = {
  },
  models: [
    {
-      id: "gemini-3.1-pro-preview",
-      name: "Gemini 3.1 Pro Preview",
+      id: "gemini-3.7-flash",
+      name: "Gemini 3.7 Flash",
      toolCalling: true,
      supportsVision: true,
    },
    {
-      id: "gemini-3-flash-preview",
-      name: "Gemini 3 Flash Preview",
+      id: "gemini-3.1-pro-preview",
+      name: "Gemini 3.1 Pro Preview",
      toolCalling: true,
      supportsVision: true,
    },
@@ -40,8 +40,8 @@ export const geminiProvider: RegistryEntry = {
      supportsVision: true,
    },
    {
-      id: "gemini-3.5-flash",
-      name: "Gemini 3.5 Flash",
+      id: "gemini-3-flash-preview",
+      name: "Gemini 3 Flash Preview",
      toolCalling: true,
      supportsVision: true,
    },
--- a/open-sse/config/providers/registry/gemini/web/index.ts
+++ b/open-sse/config/providers/registry/gemini/web/index.ts
@@ -24,8 +24,8 @@ export const gemini_webProvider: RegistryEntry = {
      supportsReasoning: false,
    },
    {
-      id: "gemini-3.5-flash",
-      name: "Gemini 3.5 Flash",
+      id: "gemini-3.7-flash",
+      name: "Gemini 3.7 Flash",
      toolCalling: false,
      supportsReasoning: false,
    },
--- a/open-sse/config/providers/registry/ghe-copilot/index.ts
+++ b/open-sse/config/providers/registry/ghe-copilot/index.ts
@@ -97,14 +97,29 @@ export const gheCopilotProvider: RegistryEntry = {
      maxOutputTokens: 64000,
    },
    {
-      id: "gemini-3.5-flash",
-      name: "Gemini 3.5 Flash",
+      id: "gemini-3.7-flash",
+      name: "Gemini 3.7 Flash",
      contextLength: 1000000,
      maxOutputTokens: 64000,
    },
-    { id: "gpt-5.6-sol", name: "GPT-5.6 Sol", targetFormat: "openai-responses", maxOutputTokens: 128000 },
-    { id: "gpt-5.6-terra", name: "GPT-5.6 Terra", targetFormat: "openai-responses", maxOutputTokens: 128000 },
-    { id: "gpt-5.6-luna", name: "GPT-5.6 Luna", targetFormat: "openai-responses", maxOutputTokens: 128000 },
+    {
+      id: "gpt-5.6-sol",
+      name: "GPT-5.6 Sol",
+      targetFormat: "openai-responses",
+      maxOutputTokens: 128000,
+    },
+    {
+      id: "gpt-5.6-terra",
+      name: "GPT-5.6 Terra",
+      targetFormat: "openai-responses",
+      maxOutputTokens: 128000,
+    },
+    {
+      id: "gpt-5.6-luna",
+      name: "GPT-5.6 Luna",
+      targetFormat: "openai-responses",
+      maxOutputTokens: 128000,
+    },
    { id: "gpt-5.5", name: "GPT-5.5", ...GPT_5_5_CODEX_CAPABILITIES, maxOutputTokens: 128000 },
    {
      id: "gpt-5.4",
--- a/open-sse/config/providers/registry/github/index.ts
+++ b/open-sse/config/providers/registry/github/index.ts
@@ -117,8 +117,8 @@ export const githubProvider: RegistryEntry = {
      maxOutputTokens: 64000,
    },
    {
-      id: "gemini-3.5-flash",
-      name: "Gemini 3.5 Flash",
+      id: "gemini-3.7-flash",
+      name: "Gemini 3.7 Flash",
      contextLength: 1000000,
      maxOutputTokens: 64000,
    },
--- a/open-sse/config/providers/registry/kie/imageModels.ts
+++ b/open-sse/config/providers/registry/kie/imageModels.ts
@@ -23,9 +23,6 @@ export const KIE_IMAGE_MODELS: KieImageModelEntry[] = [
  { id: "seedream/5.0-lite-image-to-image", name: "Seedream 5.0 Lite I2I", isMarket: true },
  { id: "z-image/4.0-text-to-image", name: "Z-Image v4.0", isMarket: true },
  { id: "z-image/4.5-text-to-image", name: "Z-Image v4.5", isMarket: true },
-  { id: "google-imagen/imagen4-fast", name: "Imagen 4 Fast", isMarket: true },
-  { id: "google-imagen/imagen4-ultra", name: "Imagen 4 Ultra", isMarket: true },
-  { id: "google-imagen/imagen4", name: "Imagen 4", isMarket: true },
  { id: "google-imagen/nano-banana-2", name: "Nano Banana 2", isMarket: true },
  { id: "google-imagen/nano-banana", name: "Nano Banana", isMarket: true },
  { id: "google-imagen/nano-banana-pro", name: "Nano Banana Pro", isMarket: true },
--- a/open-sse/config/providers/registry/kie/index.ts
+++ b/open-sse/config/providers/registry/kie/index.ts
@@ -12,16 +12,15 @@ export const kieProvider: RegistryEntry = {
  models: [
    // Sweep 2026-06-19: + current flagships the kie proxy surfaces. gemini-3-pro was
    // skipped (registry already carries the newer gemini-3-1-pro).
-    { id: "claude-opus-4-8", name: "Claude 4.8 Opus" },
-    { id: "claude-opus-4-7", name: "Claude 4.7 Opus" },
-    { id: "claude-sonnet-4-6", name: "Claude 4.6 Sonnet" },
+    { id: "claude-fable-5", name: "Claude 5 Fable" },
+    { id: "claude-opus-5", name: "Claude 5 Opus" },
+    { id: "claude-sonnet-5", name: "Claude 5 Sonnet" },
    { id: "claude-haiku-4-5", name: "Claude 4.5 Haiku" },
-    { id: "gpt-5-5", name: "GPT 5.5" },
-    { id: "gpt-5-4", name: "GPT 5.4" },
-    { id: "gpt-5-2", name: "GPT 5.2" },
+    { id: "gpt-5-6-sol", name: "GPT 5.6 Sol" },
+    { id: "gpt-5-6-terra", name: "GPT 5.6 Terra" },
+    { id: "gpt-5-6-luna", name: "GPT 5.6 Luna" },
    { id: "gemini-3-1-pro", name: "Gemini 3.1 Pro" },
-    { id: "gemini-2-5-pro", name: "Gemini 2.5 Pro" },
-    { id: "gemini-3-flash", name: "Gemini 3 Flash" },
-    { id: "gemini-3-5-flash", name: "Gemini 3.5 Flash" },
+    { id: "gemini-3-7-flash", name: "Gemini 3.7 Flash" },
+    { id: "grok-4-6", name: "Grok 4.6" },
  ],
 };
--- a/open-sse/config/providers/registry/kie/models.ts
+++ b/open-sse/config/providers/registry/kie/models.ts
@@ -11,9 +11,6 @@ export const KIE_IMAGE_MODELS = [
  { id: "seedream/5.0-lite-image-to-image", name: "Seedream 5.0 Lite I2I", isMarket: true },
  { id: "z-image/4.0-text-to-image", name: "Z-Image v4.0", isMarket: true },
  { id: "z-image/4.5-text-to-image", name: "Z-Image v4.5", isMarket: true },
-  { id: "google-imagen/imagen4-fast", name: "Imagen 4 Fast", isMarket: true },
-  { id: "google-imagen/imagen4-ultra", name: "Imagen 4 Ultra", isMarket: true },
-  { id: "google-imagen/imagen4", name: "Imagen 4", isMarket: true },
  { id: "google-imagen/nano-banana-2", name: "Nano Banana 2", isMarket: true },
  { id: "google-imagen/nano-banana", name: "Nano Banana", isMarket: true },
  { id: "google-imagen/nano-banana-pro", name: "Nano Banana Pro", isMarket: true },
--- a/open-sse/config/providers/registry/kilocode/index.ts
+++ b/open-sse/config/providers/registry/kilocode/index.ts
@@ -25,19 +25,20 @@ export const kilocodeProvider: RegistryEntry = {
  },
  models: [
    { id: "openrouter/free", name: "Free Models Router" },
-    { id: "qwen/qwen3.6-plus", name: "Qwen3.6 Plus" },
-    { id: "qwen/qwen3.5-397b-a17b", name: "Qwen3.5 397B A17B" },
-    { id: "openai/gpt-5.5", name: "GPT-5.5" },
-    { id: "openai/gpt-5.4-mini", name: "GPT-5.4 Mini" },
-    { id: "anthropic/claude-opus-4.7", name: "Claude Opus 4.7" },
-    { id: "anthropic/claude-sonnet-4.6", name: "Claude Sonnet 4.6" },
+    { id: "openai/gpt-5.6-sol", name: "GPT-5.6 Sol" },
+    { id: "openai/gpt-5.6-terra", name: "GPT-5.6 Terra" },
+    { id: "openai/gpt-5.6-luna", name: "GPT-5.6 Luna" },
+    { id: "anthropic/claude-opus-5", name: "Claude Opus 5" },
+    { id: "anthropic/claude-sonnet-5", name: "Claude Sonnet 5" },
    { id: "anthropic/claude-haiku-4.5", name: "Claude Haiku 4.5" },
    { id: "google/gemini-3.1-pro-preview", name: "Gemini 3.1 Pro" },
-    { id: "google/gemini-3-flash-preview", name: "Gemini 3 Flash" },
-    { id: "google/gemini-3.1-flash-lite", name: "Gemini 3.1 Flash Lite" },
-    { id: "deepseek/deepseek-v4-pro", name: "DeepSeek V4 Pro", supportsReasoning: true },
-    { id: "deepseek/deepseek-v4-flash", name: "DeepSeek V4 Flash", supportsReasoning: true },
-    { id: "moonshotai/kimi-k2.6", name: "Kimi K2.6" },
+    { id: "google/gemini-3.7-flash", name: "Gemini 3.7 Flash" },
+    { id: "google/gemini-3.5-flash-lite", name: "Gemini 3.5 Flash Lite" },
+    { id: "qwen/qwen3.8-max", name: "Qwen3.8 Max" },
+    { id: "qwen/qwen3.7-plus", name: "Qwen3.7 Plus" },
+    { id: "deepseek/deepseek-v4-pro-0813", name: "DeepSeek V4 Pro", supportsReasoning: true },
+    { id: "deepseek/deepseek-v4-flash-0731", name: "DeepSeek V4 Flash", supportsReasoning: true },
+    { id: "moonshotai/kimi-k3", name: "Kimi K3" },
  ],
  passthroughModels: true,
 };
--- a/open-sse/config/providers/registry/lmarena/directModels.ts
+++ b/open-sse/config/providers/registry/lmarena/directModels.ts
@@ -90,10 +90,10 @@ export const LMARENA_DIRECT_MODEL_ENTRIES: readonly LmarenaDirectModelEntry[] =
    category: "Text",
  },
  {
-    catalogId: "gemini-3.5-flash-high",
-    arenaId: "019f406f-fc33-7b9d-9571-7b8443bc7ca0",
-    publicName: "gemini-3.5-flash-high",
-    displayName: "gemini-3.5-flash-high",
+    catalogId: "gemini-3.6-flash",
+    arenaId: "019f90b1-c0ac-71ce-b295-487f261bf0f4",
+    publicName: "gemini-3.6-flash",
+    displayName: "gemini-3.6-flash",
    organization: "google",
    vision: true,
    category: "Text",
--- a/open-sse/config/providers/registry/opencode/zen/index.ts
+++ b/open-sse/config/providers/registry/opencode/zen/index.ts
@@ -25,55 +25,48 @@ export const opencode_zenProvider: RegistryEntry = {
      supportsReasoning: true,
      interleavedField: "reasoning_content",
    },
-    { id: "gpt-5-nano", name: "GPT 5 Nano", contextLength: 400000 },
-    { id: "gpt-5", name: "GPT 5" },
-    { id: "gpt-5-codex", name: "GPT 5 Codex" },
-    { id: "gpt-5.1", name: "GPT 5.1" },
-    { id: "gpt-5.1-codex", name: "GPT 5.1 Codex" },
-    { id: "gpt-5.1-codex-max", name: "GPT 5.1 Codex Max" },
-    { id: "gpt-5.1-codex-mini", name: "GPT 5.1 Codex Mini" },
-    { id: "gpt-5.2", name: "GPT 5.2" },
-    { id: "gpt-5.2-codex", name: "GPT 5.2 Codex" },
-    { id: "gpt-5.3-codex", name: "GPT 5.3 Codex" },
-    { id: "gpt-5.3-codex-spark", name: "GPT 5.3 Codex Spark" },
+    { id: "gpt-5.6-sol", name: "GPT 5.6 Sol" },
+    { id: "gpt-5.6-terra", name: "GPT 5.6 Terra" },
+    { id: "gpt-5.6-luna", name: "GPT 5.6 Luna" },
    { id: "gpt-5.4", name: "GPT 5.4" },
    { id: "gpt-5.4-mini", name: "GPT 5.4 Mini" },
    { id: "gpt-5.4-nano", name: "GPT 5.4 Nano" },
-    { id: "gpt-5.4-pro", name: "GPT 5.4 Pro" },
-    { id: "gpt-5.5", name: "GPT 5.5" },
-    { id: "gpt-5.5-pro", name: "GPT 5.5 Pro" },
+    { id: "gpt-5.3-codex-spark", name: "GPT 5.3 Codex Spark" },
+    { id: "gpt-5.1", name: "GPT 5.1" },

    // ── Claude ─────────────────────────────────────────────────
+    { id: "claude-fable-5", name: "Claude Fable 5" },
+    { id: "claude-opus-5", name: "Claude Opus 5" },
+    { id: "claude-sonnet-5", name: "Claude Sonnet 5" },
    { id: "claude-haiku-4-5", name: "Claude Haiku 4.5" },
-    { id: "claude-sonnet-4", name: "Claude Sonnet 4" },
-    { id: "claude-sonnet-4-5", name: "Claude Sonnet 4.5" },
-    { id: "claude-sonnet-4-6", name: "Claude Sonnet 4.6" },
-    { id: "claude-opus-4-1", name: "Claude Opus 4.1" },
-    { id: "claude-opus-4-5", name: "Claude Opus 4.5" },
-    { id: "claude-opus-4-6", name: "Claude Opus 4.6" },
-    { id: "claude-opus-4-7", name: "Claude Opus 4.7" },

    // ── Gemini ─────────────────────────────────────────────────
-    { id: "gemini-3-flash", name: "Gemini 3 Flash" },
+    { id: "gemini-3.7-flash", name: "Gemini 3.7 Flash" },
+    { id: "gemini-3.5-flash-lite", name: "Gemini 3.5 Flash Lite" },
    { id: "gemini-3.1-pro", name: "Gemini 3.1 Pro" },
-    { id: "gemini-3.5-flash", name: "Gemini 3.5 Flash" },
+    { id: "gemini-3-flash", name: "Gemini 3 Flash" },

    // ── Grok ───────────────────────────────────────────────────
    { id: "grok-build-0.1", name: "Grok Build 0.1" },
+    { id: "grok-4.6", name: "Grok 4.6" },
+
+    // ── Muse ───────────────────────────────────────────────────
+    { id: "muse-spark-1.2", name: "Muse Spark 1.2" },
+
+    // ── DeepSeek ────────────────────────────────────────────────
+    { id: "deepseek-v4-pro", name: "DeepSeek V4 Pro" },
+    { id: "deepseek-v4-flash", name: "DeepSeek V4 Flash" },

    // ── GLM / Z.AI ─────────────────────────────────────────────
-    { id: "glm-5", name: "GLM-5" },
-    { id: "glm-5.1", name: "GLM-5.1" },
+    { id: "glm-5.2", name: "GLM-5.2" },

    // ── MiniMax ────────────────────────────────────────────────
    // #3110: MiniMax M3 — frontier coding model with 1M context
    { id: "minimax-m3", name: "MiniMax M3", contextLength: 1048576, supportsVision: true },
-    { id: "minimax-m2.5", name: "MiniMax M2.5" },
-    { id: "minimax-m2.7", name: "MiniMax M2.7" },

    // ── Kimi / Moonshot ────────────────────────────────────────
-    { id: "kimi-k2.5", name: "Kimi K2.5" },
-    { id: "kimi-k2.6", name: "Kimi K2.6" },
+    { id: "kimi-k3", name: "Kimi K3" },
+    { id: "kimi-k2.7-code", name: "Kimi K2.7 Code" },

    // ── Qwen ───────────────────────────────────────────────────
    // Issue #2292: Qwen models return Claude-format SSE bodies even
@@ -86,12 +79,14 @@ export const opencode_zenProvider: RegistryEntry = {

    // ── Free Tier ──────────────────────────────────────────────
    // #6998 (2026-07-14): upstream free tier rotated — minimax-m2.5-free,
-    // nemotron-3-super-free and qwen3.6-plus-free were delisted (401). Replaced
-    // by the 4 entries below with upstream-verified limits.
+    // nemotron-3-super-free and qwen3.6-plus-free were delisted (401).
+    // 2026-08-17 sync: north-mini-code-free delisted; nemotron-3.5-lightning-free
+    // and laguna-s-2.1-free added.
    { id: "deepseek-v4-flash-free", name: "DeepSeek V4 Flash Free", supportsReasoning: true },
    { id: "mimo-v2.5-free", name: "MiMo V2.5 Free", contextLength: 200000 },
    { id: "hy3-free", name: "HY3 Free", contextLength: 200000 },
    { id: "nemotron-3-ultra-free", name: "Nemotron 3 Ultra Free", contextLength: 1000000 },
-    { id: "north-mini-code-free", name: "North Mini Code Free", contextLength: 200000 },
+    { id: "nemotron-3.5-lightning-free", name: "Nemotron 3.5 Lightning Free" },
+    { id: "laguna-s-2.1-free", name: "Laguna S 2.1 Free" },
  ],
 };
--- a/open-sse/config/providers/registry/orcarouter/index.ts
+++ b/open-sse/config/providers/registry/orcarouter/index.ts
@@ -35,8 +35,8 @@ export const orcarouterProvider: RegistryEntry = {
      maxOutputTokens: 128000,
    },
    {
-      id: "google/gemini-3.5-flash",
-      name: "Gemini 3.5 Flash",
+      id: "google/gemini-3.6-flash",
+      name: "Gemini 3.6 Flash",
      toolCalling: true,
      supportsReasoning: true,
      supportsVision: true,
--- a/open-sse/config/providers/registry/perplexity/web/index.ts
+++ b/open-sse/config/providers/registry/perplexity/web/index.ts
@@ -13,12 +13,12 @@ export const perplexity_webProvider: RegistryEntry = {
    { id: "pplx-sonar", name: "Sonar 2 (via Perplexity)", toolCalling: false },
    { id: "pplx-gpt-5.6-terra", name: "GPT-5.6 Terra (via Perplexity)", toolCalling: false },
    { id: "pplx-gpt-5.6-sol", name: "GPT-5.6 Sol (via Perplexity)", toolCalling: false },
-    { id: "pplx-gemini", name: "Gemini 3.1 Pro (via Perplexity)", toolCalling: false },
+    { id: "pplx-gemini", name: "Gemini 3.7 Flash (via Perplexity)", toolCalling: false },
    { id: "pplx-sonnet", name: "Claude Sonnet 5.0 (via Perplexity)", toolCalling: false },
    { id: "pplx-opus", name: "Claude Opus 5.0 (via Perplexity)", toolCalling: false },
    { id: "pplx-glm", name: "GLM-5.2 (via Perplexity)", toolCalling: false },
-    { id: "pplx-kimi", name: "Kimi K2.6 (via Perplexity)", toolCalling: false },
-    { id: "pplx-grok-4.5", name: "Grok 4.5 (via Perplexity)", toolCalling: false },
+    { id: "pplx-kimi", name: "Kimi K3 (via Perplexity)", toolCalling: false },
+    { id: "pplx-grok-4.6", name: "Grok 4.6 (via Perplexity)", toolCalling: false },
    { id: "pplx-nemotron", name: "Nemotron 3 Ultra (via Perplexity)", toolCalling: false },
  ],
 };
--- a/open-sse/config/providers/registry/raycast/index.ts
+++ b/open-sse/config/providers/registry/raycast/index.ts
@@ -19,15 +19,43 @@ export const raycastProvider: RegistryEntry = {
  authHeader: "bearer",
  defaultContextLength: 128000,
  models: [
-    { id: "openai-gpt-5-mini", name: "GPT-5 Mini" },
-    { id: "openai-gpt-4o-mini", name: "GPT-4o Mini" },
-    { id: "anthropic-claude-sonnet-4-6", name: "Claude Sonnet 4.6" },
-    { id: "google-gemini-2.5-flash", name: "Gemini 2.5 Flash" },
-    { id: "raycast-ray1", name: "Ray1" },
-    { id: "raycast-ray1-mini", name: "Ray1 Mini" },
-    { id: "perplexity-sonar", name: "Sonar" },
+    //GPT
+    { id: "openai-gpt-5.6-sol", name: "GPT-5.6 Sol" },
+    { id: "openai-gpt-5.6-terra", name: "GPT-5.6 Terra" },
+    { id: "openai-gpt-5.6-luna", name: "GPT-5.6 Luna" },
+    //Claude
+    { id: "anthropic-claude-opus-5", name: "Claude Opus 5" },
+    { id: "anthropic-claude-sonnet-5", name: "Claude Sonnet 5" },
+    { id: "anthropic-claude-4-5-haiku-reasoning", name: "Claude 4.5 Haiku Reasoning" },
+    { id: "anthropic-claude-4-5-haiku", name: "Claude 4.5 Haiku" },
+    //Gemini
+    { id: "google-gemini-3.1-pro", name: "Gemini 3.1 Pro" },
+    { id: "google-gemini-3.7-flash", name: "Gemini 3.7 Flash" },
+    { id: "google-gemini-3.5-flash-lite", name: "Gemini 3.5 Flash Lite" },
+    //Perplexity
+    { id: "perplexity-sonar-reasoning-pro", name: "Sonar Reasoning Pro" },
    { id: "perplexity-sonar-pro", name: "Sonar Pro" },
+    { id: "perplexity-sonar", name: "Sonar" },
+    //Mistral
+    { id: "mistral-mistral-large-latest", name: "Mistral Large" },
+    { id: "mistral-mistral-medium-latest", name: "Mistral Medium" },
+    { id: "mistral-mistral-small-latest", name: "Mistral Small" },
+    { id: "mistral-codestral-latest", name: "Codestral" },
    { id: "mistral-open-mistral-nemo", name: "Mistral Nemo" },
-    { id: "xai-grok-3-mini", name: "Grok 3 Mini" },
+    //Grok
+    { id: "xai-grok-4.6", name: "Grok 4.6" },
+    //Opensource
+    { id: "gateway-alibaba/qwen3.8-max", name: "Qwen 3.8 Max" },
+    { id: "gateway-moonshotai/kimi-k3", name: "Kimi K3" },
+    { id: "baseten-deepseek-ai/DeepSeek-V4-Pro", name: "DeepSeek V4 Pro" },
+    { id: "gateway-deepseek/deepseek-v4-flash", name: "DeepSeek V4 Flash" },
+    { id: "baseten-zai-org/GLM-5.2", name: "GLM 5.2" },
+    { id: "gateway-thinkingmachines/inkling-1.0", name: "Inkling 1.0" },
+    { id: "gateway-google/gemma-4-31b-it", name: "Gemma 4 31B" },
+    { id: "groq-openai/gpt-oss-120b", name: "GPT-OSS 120B" },
+    { id: "groq-openai/gpt-oss-20b", name: "GPT-OSS 20B" },
+    { id: "groq-qwen/qwen3-32b", name: "Qwen 3 32B" },
+    { id: "groq-llama-3.3-70b-versatile", name: "LLaMA 3.3 70B" },
+    { id: "groq-llama-3.1-8b-instant", name: "LLaMA 3.1 8B" },
  ],
 };
--- a/open-sse/config/providers/registry/tinycms/index.ts
+++ b/open-sse/config/providers/registry/tinycms/index.ts
@@ -18,22 +18,28 @@ export const tinycmsProvider: RegistryEntry = {
  authType: "apikey",
  authHeader: "uuid",
  models: [
-    { id: "gpt-5-free", name: "GPT 5 Free" },
-    { id: "gpt-5.3-free", name: "GPT 5.3 Free (Multimodal/Vision)" },
-    { id: "gpt-5.3-thinking-free", name: "GPT 5.3 Thinking Free", supportsReasoning: true },
+    { id: "claude-fable-5", name: "Claude Fable 5" },
+    { id: "claude-opus-5", name: "Claude Opus 5" },
+    { id: "claude-sonnet-5", name: "Claude Sonnet 5" },
+    { id: "gpt-5.6-sol", name: "GPT 5.6 Sol" },
+    { id: "gpt-5.6-luna", name: "GPT 5.6 Luna" },
+    { id: "gpt-5.5", name: "GPT 5.5" },
    { id: "gpt-5.4-mini", name: "GPT 5.4 Mini" },
    { id: "gpt-5.4-nano", name: "GPT 5.4 Nano" },
-    { id: "gpt-5-nano", name: "GPT 5 Nano" },
-    { id: "gemini-3.5-flash", name: "Gemini 3.5 Flash" },
-    { id: "gemini-3-pro-preview", name: "Gemini 3 Pro Preview" },
-    { id: "gemini-3.1-flash-lite-preview", name: "Gemini 3.1 Flash Lite Preview" },
-    { id: "grok-4.20-fast", name: "Grok 4.20 Fast" },
-    { id: "grok-4.20", name: "Grok 4.20" },
-    { id: "grok-imagine", name: "Grok Imagine (Image Gen)" },
+    { id: "gpt-5.3-thinking-free", name: "GPT 5.3 Thinking Free", supportsReasoning: true },
+    { id: "gpt-5.3-free", name: "GPT 5.3 Free (Multimodal/Vision)" },
+    { id: "gpt-oss-120b", name: "GPT-OSS 120B" },
+    { id: "gemini-3.6-flash", name: "Gemini 3.6 Flash" },
+    { id: "gemini-3.1-pro-preview", name: "Gemini 3.1 Pro" },
+    { id: "gemini-3.1-flash-lite-preview", name: "Gemini 3.1 Flash Lite" },
+    { id: "grok-4.5", name: "Grok 4.5" },
+    { id: "deepseek-v4-pro", name: "DeepSeek V4 Pro" },
    { id: "deepseek-v4-flash", name: "DeepSeek V4 Flash" },
-    { id: "claude-sonnet-5", name: "Claude Sonnet 5" },
-    { id: "gpt-image-2", name: "GPT Image 2 (Image Gen)" },
+    { id: "kimi-k3", name: "Kimi K3" },
+    { id: "glm-5.2", name: "GLM 5.2" },
    { id: "qwen3.6-plus", name: "Qwen 3.6 Plus" },
+    { id: "mimo-v2.5-pro", name: "Mimo V2.5 Pro" },
+    { id: "mimo-v2.5", name: "Mimo V2.5" },
  ],
 };

--- a/open-sse/executors/antigravityOutputCap.ts
+++ b/open-sse/executors/antigravityOutputCap.ts
@@ -1,4 +1,5 @@
 import { getExplicitModelOutputCap } from "@/lib/modelCapabilities";
+import { isDiscoverableAntigravityModelId } from "../config/antigravityModelAliases";

 /**
 * Fallback ceiling on `generationConfig.maxOutputTokens` for Antigravity
@@ -13,7 +14,7 @@ import { getExplicitModelOutputCap } from "@/lib/modelCapabilities";
 * Both of those models are catalogue-known today, so neither one reaches this
 * constant anymore: they get their own declared limit via
 * `resolveAntigravityOutputCap` (65536 and 65535, respectively). The higher
- * limit holds against the live upstream. A gemini-3.6-flash-high request came
+ * limit holds against the live upstream. A gemini-3.7-flash-high request came
 * back with completion_tokens 16754 and finish_reason "stop", which exceeds
 * 16384 on its own and so cannot be an artifact of thinking-token accounting.
 *
@@ -38,6 +39,10 @@ export const MAX_ANTIGRAVITY_OUTPUT_TOKENS = 16384;
 export function resolveAntigravityOutputCap(modelId: string | null | undefined): number {
  const id = typeof modelId === "string" ? modelId.trim() : "";
  if (!id) return MAX_ANTIGRAVITY_OUTPUT_TOKENS;
+  // MODEL_SPECS is provider-neutral: other providers may continue serving old
+  // Gemini 3.5/3.6 ids after Antigravity retires them. Do not let those shared
+  // specs make a retired Antigravity id look active on this provider path.
+  if (!isDiscoverableAntigravityModelId(id)) return MAX_ANTIGRAVITY_OUTPUT_TOKENS;
  try {
    const declared = getExplicitModelOutputCap({ provider: "antigravity", model: id });
    return typeof declared === "number" && Number.isFinite(declared) && declared > 0
--- a/open-sse/executors/perplexity-web/protocol.ts
+++ b/open-sse/executors/perplexity-web/protocol.ts
@@ -69,24 +69,26 @@ export const MODEL_MAP: Record<string, [string, string]> = {
  "pplx-sonar": ["copilot", "turbo"],
  "pplx-gpt-5.6-terra": ["copilot", "gpt56_terra"],
  "pplx-gpt-5.6-sol": ["copilot", "gpt56_sol"],
-  "pplx-gemini": ["copilot", "gemini31pro_high"],
+  "pplx-gemini": ["copilot", "gemini37flash"],
  "pplx-sonnet": ["copilot", "claude50sonnet"],
  // Perplexity's catalog moved Opus to 5.0; claude48opus is still accepted but
  // answers from the older model.
  "pplx-opus": ["copilot", "claude50opus"],
  "pplx-glm": ["copilot", "glm_5_2"],
-  "pplx-kimi": ["copilot", "kimik26instant"],
-  "pplx-grok-4.5": ["copilot", "grok45low"],
+  // The current Kimi K3 catalog entry only exposes its reasoning model.
+  "pplx-kimi": ["copilot", "kimik3thinking"],
+  "pplx-grok-4.6": ["copilot", "grok46low"],
  "pplx-nemotron": ["copilot", "nv_nemotron_3_ultra"],
 };

 export const THINKING_MAP: Record<string, string> = {
  "pplx-gpt-5.6-terra": "gpt56_terra_thinking",
  "pplx-gpt-5.6-sol": "gpt56_sol_thinking",
+  "pplx-gemini": "gemini37flashthinking",
  "pplx-sonnet": "claude50sonnetthinking",
  "pplx-opus": "claude50opusthinking",
-  "pplx-kimi": "kimik26thinking",
-  "pplx-grok-4.5": "grok45medium",
+  "pplx-kimi": "kimik3thinking",
+  "pplx-grok-4.6": "grok46medium",
 };

 export const CITATION_RE = /\[\d+\]/g;
--- a/open-sse/handlers/imageGeneration.ts
+++ b/open-sse/handlers/imageGeneration.ts
@@ -41,7 +41,6 @@ import { handleHyperbolicImageGeneration } from "./imageGeneration/providers/hyp
 import { handleHuggingFaceImageGeneration } from "./imageGeneration/providers/huggingface.ts";
 import { handleComfyUIImageGeneration } from "./imageGeneration/providers/comfyUI.ts";
 import { handleImagen3ImageGeneration } from "./imageGeneration/providers/imagen3.ts";
-import { handleGoogleImagenGeneration } from "./imageGeneration/providers/googleImagen.ts";
 import { handleIdeogramImageGeneration } from "./imageGeneration/providers/ideogram.ts";
 import { handleHaiperImageGeneration } from "./imageGeneration/providers/haiper.ts";
 import { handleLeonardoImageGeneration } from "./imageGeneration/providers/leonardo.ts";
@@ -383,17 +382,6 @@ export async function handleImageGeneration({
    });
  }

-  if (providerConfig.format === "google-imagen") {
-    return handleGoogleImagenGeneration({
-      model,
-      provider,
-      providerConfig,
-      body,
-      credentials,
-      log,
-    });
-  }
-
  if (providerConfig.format === "hyperbolic") {
    return handleHyperbolicImageGeneration({
      model,
--- a/open-sse/handlers/imageGeneration/providers/googleImagen.ts
+++ b/open-sse/handlers/imageGeneration/providers/googleImagen.ts
@@ -1,147 +0,0 @@
-// Google AI Studio (Gemini API) Imagen image generation.
-//
-// Unlike the antigravity "gemini-image" format (which wraps generateContent in a
-// Cloud Code envelope), the Imagen family on generativelanguage.googleapis.com uses
-// the dedicated ":predict" endpoint with an instances/parameters body and returns
-// base64 image bytes under `predictions[].bytesBase64Encoded`.
-//
-// Docs: https://ai.google.dev/gemini-api/docs/imagen  (Imagen requires a billing-
-// enabled Google project; free-tier keys get 403 / quota 0.)
-
-import { saveCallLog } from "@/lib/usageDb";
-import { mapImageSize } from "../../../translator/image/sizeMapper.ts";
-import { sanitizeErrorMessage } from "../../../utils/error.ts";
-
-// Only the Imagen family routes through :predict. Other gemini image models
-// (gemini-*-flash-image / nano-banana) use generateContent and belong on the chat
-// route, so they must not be dispatched here.
-export function isImagenModel(model) {
-  return /^imagen-/i.test(String(model || ""));
-}
-
-/**
- * Build the Imagen :predict request body from an OpenAI-style image request.
- * Pure — no I/O — so it can be unit-tested without live credentials.
- */
-export function buildImagenPredictBody(body) {
-  const prompt = typeof body?.prompt === "string" ? body.prompt : String(body?.prompt ?? "");
-  const n = Number(body?.n);
-  const sampleCount = Number.isFinite(n) && n > 0 ? Math.min(Math.floor(n), 4) : 1;
-  return {
-    instances: [{ prompt }],
-    parameters: {
-      sampleCount,
-      aspectRatio: mapImageSize(body?.aspect_ratio || body?.size),
-    },
-  };
-}
-
-/**
- * Normalize an Imagen :predict response into the OpenAI image-generation shape
- * ({ created, data: [{ b64_json, revised_prompt }] }). Pure — unit-testable.
- */
-export function parseImagenPredictResponse(data, prompt) {
-  const predictions = Array.isArray(data?.predictions) ? data.predictions : [];
-  const images = [];
-  for (const p of predictions) {
-    const b64 = p?.bytesBase64Encoded ?? p?.b64_json ?? p?.image ?? null;
-    if (typeof b64 === "string" && b64.length > 0) {
-      images.push({ b64_json: b64, revised_prompt: prompt });
-    }
-  }
-  return { created: Math.floor(Date.now() / 1000), data: images };
-}
-
-export async function handleGoogleImagenGeneration({
-  model,
-  provider,
-  providerConfig,
-  body,
-  credentials,
-  log,
-}) {
-  const startTime = Date.now();
-  const token = credentials?.apiKey || credentials?.accessToken || "";
-  const prompt = typeof body.prompt === "string" ? body.prompt : String(body.prompt ?? "");
-
-  if (!isImagenModel(model)) {
-    return {
-      success: false,
-      status: 400,
-      error: `Model ${model} is not an Imagen model. Gemini flash-image models route through /v1/chat/completions, not /v1/images/generations.`,
-    };
-  }
-
-  const upstreamBody = buildImagenPredictBody(body);
-  // baseUrl is https://generativelanguage.googleapis.com/v1beta/models
-  const url = `${providerConfig.baseUrl.replace(/\/$/, "")}/${model}:predict`;
-
-  if (log) {
-    log.info(
-      "IMAGE",
-      `${provider}/${model} (google-imagen) | prompt: "${prompt.slice(0, 60)}..." | aspectRatio: ${upstreamBody.parameters.aspectRatio}`
-    );
-  }
-
-  try {
-    const response = await fetch(url, {
-      method: "POST",
-      headers: {
-        "Content-Type": "application/json",
-        // Key travels in the header, never the URL, so it stays out of logs.
-        "x-goog-api-key": token,
-      },
-      body: JSON.stringify(upstreamBody),
-    });
-
-    if (!response.ok) {
-      const errorText = await response.text();
-      const safeError = sanitizeErrorMessage(errorText);
-      if (log) log.error("IMAGE", `${provider} error ${response.status}: ${safeError.slice(0, 200)}`);
-
-      saveCallLog({
-        method: "POST",
-        path: "/v1/images/generations",
-        status: response.status,
-        model: `${provider}/${model}`,
-        provider,
-        duration: Date.now() - startTime,
-        error: safeError.slice(0, 500),
-      }).catch(() => {});
-
-      return { success: false, status: response.status, error: safeError };
-    }
-
-    const data = await response.json();
-    const normalized = parseImagenPredictResponse(data, prompt);
-
-    saveCallLog({
-      method: "POST",
-      path: "/v1/images/generations",
-      status: 200,
-      model: `${provider}/${model}`,
-      provider,
-      duration: Date.now() - startTime,
-      responseBody: { images_count: normalized.data.length },
-    }).catch(() => {});
-
-    return { success: true, data: normalized };
-  } catch (err) {
-    const errMsg = err instanceof Error ? err.message : String(err);
-    if (log) log.error("IMAGE", `${provider} fetch error: ${errMsg}`);
-    saveCallLog({
-      method: "POST",
-      path: "/v1/images/generations",
-      status: 502,
-      model: `${provider}/${model}`,
-      provider,
-      duration: Date.now() - startTime,
-      error: errMsg,
-    }).catch(() => {});
-    return {
-      success: false,
-      status: 502,
-      error: `Image provider error: ${sanitizeErrorMessage(errMsg)}`,
-    };
-  }
-}
--- a/open-sse/services/tests/antigravity-quota-family.test.ts
+++ b/open-sse/services/tests/antigravity-quota-family.test.ts
@@ -22,24 +22,24 @@ describe("Antigravity account quota-family cooldown", () => {
  });

  it("maps Gemini variants to Gemini family and Claude/Cloud variants to Claude family", () => {
-    expect(getAntigravityQuotaFamily("gemini-3.5-flash-medium")).toBe("gemini");
-    expect(getAntigravityQuotaFamily("google/gemini-3.5-flash-low")).toBe("gemini");
-    expect(getAntigravityQuotaFamily("agy/gemini-3.5-flash-medium")).toBe("gemini");
+    expect(getAntigravityQuotaFamily("gemini-3.7-flash-medium")).toBe("gemini");
+    expect(getAntigravityQuotaFamily("google/gemini-3.7-flash-low")).toBe("gemini");
+    expect(getAntigravityQuotaFamily("agy/gemini-3.7-flash-medium")).toBe("gemini");
    expect(getAntigravityQuotaFamily("claude-sonnet-4")).toBe("claude");
    expect(getAntigravityQuotaFamily("cloud/claude-opus-4")).toBe("claude");
    expect(getAntigravityQuotaFamily("some-new-model")).toBe("other");
  });

  it("uses family-scoped lock key for Antigravity but preserves exact-model scope elsewhere", () => {
-    expect(getQuotaScopedModelForProvider("antigravity", "gemini-3.5-flash-medium")).toBe(
+    expect(getQuotaScopedModelForProvider("antigravity", "gemini-3.7-flash-medium")).toBe(
      "family:gemini"
    );
-    expect(getQuotaScopedModelForProvider("agy", "gemini-3.5-flash-medium")).toBe("family:gemini");
-    expect(getQuotaScopedModelForProvider(provider, "gemini-3.5-flash-low")).toBe("family:gemini");
+    expect(getQuotaScopedModelForProvider("agy", "gemini-3.7-flash-medium")).toBe("family:gemini");
+    expect(getQuotaScopedModelForProvider(provider, "gemini-3.7-flash-low")).toBe("family:gemini");
    expect(getQuotaScopedModelForProvider(provider, "claude-sonnet-4")).toBe("family:claude");
    expect(getQuotaScopedModelForProvider(provider, "unknown-model")).toBe("unknown-model");
-    expect(getQuotaScopedModelForProvider("openai", "gemini-3.5-flash-medium")).toBe(
-      "gemini-3.5-flash-medium"
+    expect(getQuotaScopedModelForProvider("openai", "gemini-3.7-flash-medium")).toBe(
+      "gemini-3.7-flash-medium"
    );
  });

@@ -47,7 +47,7 @@ describe("Antigravity account quota-family cooldown", () => {
    recordModelLockoutFailure(
      provider,
      "account-a",
-      "gemini-3.5-flash-medium",
+      "gemini-3.7-flash-medium",
      "rate_limited",
      429,
      60_000,
@@ -55,10 +55,10 @@ describe("Antigravity account quota-family cooldown", () => {
      { maxCooldownMs: 300_000 }
    );

-    expect(isModelLocked(provider, "account-a", "gemini-3.5-flash-medium")).toBe(true);
-    expect(isModelLocked(provider, "account-a", "gemini-3.5-flash-low")).toBe(true);
+    expect(isModelLocked(provider, "account-a", "gemini-3.7-flash-medium")).toBe(true);
+    expect(isModelLocked(provider, "account-a", "gemini-3.7-flash-low")).toBe(true);
    expect(isModelLocked(provider, "account-a", "claude-sonnet-4")).toBe(false);
-    expect(isModelLocked(provider, "account-b", "gemini-3.5-flash-low")).toBe(false);
+    expect(isModelLocked(provider, "account-b", "gemini-3.7-flash-low")).toBe(false);
  });

  it("keeps Claude/Cloud family distinct from Gemini", () => {
@@ -74,7 +74,7 @@ describe("Antigravity account quota-family cooldown", () => {
    );

    expect(isModelLocked(provider, "account-a", "cloud/claude-opus-4")).toBe(true);
-    expect(isModelLocked(provider, "account-a", "gemini-3.5-flash-low")).toBe(false);
+    expect(isModelLocked(provider, "account-a", "gemini-3.7-flash-low")).toBe(false);
  });

  it("can isolate a confirmed Antigravity quota exhaustion to one exact model", () => {
@@ -88,7 +88,7 @@ describe("Antigravity account quota-family cooldown", () => {

    expect(isModelLocked(provider, "account-a", "claude-opus-4-6-thinking")).toBe(true);
    expect(isModelLocked(provider, "account-a", "claude-sonnet-4-6-thinking")).toBe(false);
-    expect(isModelLocked(provider, "account-a", "gemini-3.5-flash-medium")).toBe(false);
+    expect(isModelLocked(provider, "account-a", "gemini-3.7-flash-medium")).toBe(false);

    expect(clearModelLock(provider, "account-a", "claude-opus-4-6-thinking")).toBe(true);
    expect(isModelLocked(provider, "account-a", "claude-opus-4-6-thinking")).toBe(false);
@@ -113,7 +113,7 @@ describe("Antigravity account quota-family cooldown", () => {
    const upstream = recordModelLockoutFailure(
      provider,
      "account-a",
-      "gemini-3.5-flash-medium",
+      "gemini-3.7-flash-medium",
      "rate_limited",
      429,
      1_000,
@@ -122,13 +122,13 @@ describe("Antigravity account quota-family cooldown", () => {
    );
    expect(upstream.cooldownMs).toBe(123_000);
    expect(
-      getModelLockoutInfo(provider, "account-a", "gemini-3.5-flash-low")?.remainingMs
+      getModelLockoutInfo(provider, "account-a", "gemini-3.7-flash-low")?.remainingMs
    ).toBeGreaterThan(100_000);

    const inferred = recordModelLockoutFailure(
      provider,
      "account-b",
-      "gemini-3.5-flash-medium",
+      "gemini-3.7-flash-medium",
      "rate_limited",
      429,
      1_000,
--- a/open-sse/services/conolModels.ts
+++ b/open-sse/services/conolModels.ts
@@ -49,7 +49,6 @@ const FALLBACK_MODEL_SEEDS: FallbackModelSeed[] = [
  { id: "claude-opus-5", vision: true, efforts: EFFORTS_XHIGH },
  { id: "claude-opus-4-8", vision: true, efforts: EFFORTS_XHIGH },
  { id: "claude-fable-5", vision: true, efforts: EFFORTS_XHIGH },
-  { id: "claude-opus-4-7", vision: true, efforts: EFFORTS_XHIGH },
  { id: "claude-sonnet-5", vision: true, efforts: EFFORTS_NO_XHIGH },
  { id: "claude-sonnet-4-6", vision: true, efforts: EFFORTS_NO_XHIGH },
  { id: "claude-haiku-4-5", vision: true, efforts: EFFORTS_STANDARD },
@@ -61,7 +60,6 @@ const FALLBACK_MODEL_SEEDS: FallbackModelSeed[] = [
  { id: "deepseek/deepseek-v4-pro", vision: false, efforts: EFFORTS_HIGH_ONLY },
  { id: "openrouter/fusion", vision: false, efforts: [] },
  { id: "z-ai/glm-5.2", vision: false, efforts: EFFORTS_STANDARD },
-  { id: "z-ai/glm-5.1", vision: false, efforts: EFFORTS_STANDARD },
  { id: "tencent/hy3", vision: false, efforts: EFFORTS_STANDARD },
  { id: "moonshotai/kimi-k3", vision: true, efforts: EFFORTS_STANDARD },
  { id: "moonshotai/kimi-k2.7-code", vision: true, efforts: EFFORTS_STANDARD },
@@ -69,7 +67,7 @@ const FALLBACK_MODEL_SEEDS: FallbackModelSeed[] = [
  { id: "qwen/qwen3.7-max", vision: false, efforts: EFFORTS_STANDARD },
  { id: "minimax/minimax-m3", vision: true, efforts: EFFORTS_STANDARD },
  { id: "stepfun/step-3.7-flash", vision: true, efforts: EFFORTS_STANDARD },
-  { id: "google/gemini-3.5-flash", vision: true, efforts: EFFORTS_STANDARD },
+  { id: "google/gemini-3.7-flash", vision: true, efforts: EFFORTS_STANDARD },
  { id: "google/gemini-3.1-pro-preview", vision: true, efforts: EFFORTS_STANDARD },
  { id: "google/gemini-3.1-flash-lite", vision: true, efforts: EFFORTS_STANDARD },
  { id: "x-ai/grok-4.3", vision: true, efforts: EFFORTS_STANDARD },
--- a/open-sse/services/githubCopilotModels.ts
+++ b/open-sse/services/githubCopilotModels.ts
@@ -32,7 +32,7 @@ export const GITHUB_COPILOT_MODEL_ALLOWLIST = [
  "claude-sonnet-4.5",
  "claude-haiku-4.5",
  "gemini-3.1-pro-preview",
-  "gemini-3.5-flash",
+  "gemini-3.7-flash",
  "gpt-5.6-sol",
  "gpt-5.6-terra",
  "gpt-5.6-luna",
--- a/open-sse/services/notionWebFallbackModels.ts
+++ b/open-sse/services/notionWebFallbackModels.ts
@@ -30,84 +30,96 @@ export type NotionDiscoveredModel = {
 */
 export const NOTION_WEB_FALLBACK_MODELS: NotionDiscoveredModel[] = [
  { id: "notion-ai", name: "Notion AI (default)", owned_by: "notion" },
-  { id: "gpt-5.6-sol", name: "GPT-5.6 Sol", owned_by: "openai", notionCodename: "orange-mousse" },
+  {
+    id: "gpt-5.6-sol",
+    name: "GPT-5.6 Sol",
+    owned_by: "openai",
+    supportsReasoning: true,
+    notionCodename: "orange-mousse",
+  },
  {
    id: "gpt-5.6-terra",
    name: "GPT-5.6 Terra",
    owned_by: "openai",
+    supportsReasoning: true,
    notionCodename: "orchid-muffin",
  },
  {
    id: "gpt-5.6-luna",
    name: "GPT-5.6 Luna",
    owned_by: "openai",
+    supportsReasoning: true,
    notionCodename: "olive-jellyroll",
  },
-  { id: "gpt-5.2", name: "GPT-5.2", owned_by: "openai", notionCodename: "oatmeal-cookie" },
-  { id: "gpt-5.4", name: "GPT-5.4", owned_by: "openai", notionCodename: "oval-kumquat-medium" },
-  { id: "gpt-5.5", name: "GPT-5.5", owned_by: "openai", notionCodename: "opal-quince-medium" },
  {
    id: "gpt-5.4-mini",
    name: "GPT-5.4 Mini",
    owned_by: "openai",
+    supportsReasoning: true,
    notionCodename: "oregon-grape-medium",
  },
  {
    id: "gpt-5.4-nano",
    name: "GPT-5.4 Nano",
    owned_by: "openai",
+    supportsReasoning: true,
    notionCodename: "otaheite-apple-medium",
  },
  {
-    id: "gemini-3.5-flash",
-    name: "Gemini 3.5 Flash",
+    id: "gemini-3.7-flash",
+    name: "Gemini 3.7 Flash",
    owned_by: "gemini",
-    notionCodename: "vertex-gemini-3.5-flash",
-  },
-  {
-    id: "gemini-3-flash",
-    name: "Gemini 3 Flash",
-    owned_by: "gemini",
-    notionCodename: "gingerbread",
+    supportsReasoning: true,
+    notionCodename: "grapefruit-zeppole",
  },
  {
    id: "gemini-3.1-pro",
    name: "Gemini 3.1 Pro",
    owned_by: "gemini",
+    supportsReasoning: true,
    notionCodename: "galette-medium-thinking",
  },
  {
-    id: "sonnet-4.6",
-    name: "Sonnet 4.6",
+    id: "fable-5",
+    name: "Claude Fable 5",
    owned_by: "anthropic",
-    notionCodename: "almond-croissant-low",
-  },
-  { id: "sonnet-5", name: "Sonnet 5", owned_by: "anthropic", notionCodename: "angel-cake-high" },
-  {
-    id: "opus-4.6",
-    name: "Opus 4.6",
-    owned_by: "anthropic",
-    notionCodename: "avocado-froyo-medium",
+    supportsReasoning: true,
+    disabled: true,
+    notionCodename: "acai-budino-high",
  },
  {
-    id: "opus-4.7",
-    name: "Opus 4.7",
+    id: "opus-5",
+    name: "Claude Opus 5",
    owned_by: "anthropic",
-    notionCodename: "apricot-sorbet-high",
+    supportsReasoning: true,
+    notionCodename: "agave-flan",
+  },
+  {
+    id: "sonnet-5",
+    name: "Claude Sonnet 5",
+    owned_by: "anthropic",
+    supportsReasoning: true,
+    notionCodename: "angel-cake-high",
  },
-  { id: "opus-4.8", name: "Opus 4.8", owned_by: "anthropic", notionCodename: "ambrosia-tart-high" },
  {
    id: "haiku-4.5",
-    name: "Haiku 4.5",
+    name: "Claude Haiku 4.5",
    owned_by: "anthropic",
    notionCodename: "anthropic-haiku-4.5",
  },
-  { id: "fable-5", name: "Fable 5", owned_by: "anthropic", notionCodename: "acai-budino-high" },
  {
-    id: "kimi-k2.6",
-    name: "Kimi K2.6",
+    id: "grok-4.6",
+    name: "Grok 4.6",
+    owned_by: "xai",
+    supportsReasoning: true,
+    notionCodename: "soursop-shortcake",
+  },
+  {
+    id: "kimi-k3",
+    name: "Kimi K3",
    owned_by: "mystery",
-    notionCodename: "fireworks-kimi-k2.6",
+    supportsReasoning: true,
+    notionCodename: "fireworks-kimi-k3",
  },
  {
    id: "kimi-k2.7-code",
@@ -119,16 +131,14 @@ export const NOTION_WEB_FALLBACK_MODELS: NotionDiscoveredModel[] = [
    id: "deepseek-v4-pro",
    name: "DeepSeek V4 Pro",
    owned_by: "mystery",
+    supportsReasoning: true,
    notionCodename: "baseten-deepseek-v4-pro",
  },
-  { id: "glm-5.2", name: "GLM 5.2", owned_by: "mystery", notionCodename: "baseten-glm-5.2" },
-  { id: "grok-4.3", name: "Grok 4.3", owned_by: "xai", notionCodename: "xigua-mochi-medium" },
-  { id: "grok-4.5", name: "Grok 4.5", owned_by: "xai", notionCodename: "strawberry-whoopiepie" },
  {
-    id: "grok-build-0.1",
-    name: "Grok Build 0.1",
-    owned_by: "xai",
-    notionCodename: "xinomavro-cake",
+    id: "glm-5.2",
+    name: "GLM 5.2",
+    owned_by: "mystery",
+    supportsReasoning: true,
+    notionCodename: "baseten-glm-5.2",
  },
 ];
-
--- a/open-sse/services/promptqlModels.ts
+++ b/open-sse/services/promptqlModels.ts
@@ -2,7 +2,7 @@
 * PromptQL (prompt.ql.app) model catalog helpers.
 *
 * Live catalog: GraphQL `FetchLlmConfigs` against the playground Hasura endpoint.
- * Fallback: static seed captured 2026-07-20 (display_label / model_reference / model_id).
+ * Fallback: static seed captured 2026-08-17 (display_label / model_reference / model_id).
 */

 export interface PromptQlModel {
@@ -21,95 +21,102 @@ export interface PromptQlModel {
 /** Offline seed when discovery fails (from live FetchLlmConfigs capture). */
 export const PROMPTQL_FALLBACK_MODELS: PromptQlModel[] = [
  {
-    id: "vertex-claude-fable-5",
+    id: "bedrock-claude-fable-5",
    name: "Claude Fable 5",
-    configId: "967e6517-1d6b-4e22-82fb-3463bab239c4",
-    modelId: "anthropic/claude-fable-5",
+    configId: "c47a1e57-2fca-4cfe-913a-5fb821079f50",
+    modelId: "us.anthropic.claude-fable-5",
+    supportsVision: true,
  },
  {
-    id: "bedrock-claude-opus-4-8",
-    name: "Claude Opus 4.8",
-    configId: "e97e7f50-9e4a-4685-bc14-1854f1f79782",
-    modelId: "us.anthropic.claude-opus-4-8",
+    id: "bedrock-claude-opus-5",
+    name: "Claude Opus 5",
+    configId: "8aed42aa-f7c8-48f9-8238-5046bbc0f4f7",
+    modelId: "us.anthropic.claude-opus-5",
+    supportsVision: true,
  },
  {
    id: "bedrock-claude-sonnet-4-5",
    name: "Claude Sonnet 4.5",
-    configId: "48105d83-9a45-4ec6-8b58-f3cf44094f92",
+    configId: "0abcbc61-dbef-4958-96b9-e0cde7e3ad8f",
    modelId: "us.anthropic.claude-sonnet-4-5-20250929-v1:0",
+    supportsVision: true,
  },
  {
-    id: "deepseek-v4-pro",
-    name: "DeepSeek V4 Pro",
-    configId: "5a23af33-b31b-4215-892c-20ef633a8848",
-    modelId: "accounts/fireworks/models/deepseek-v4-pro",
+    id: "deepseek-v4-pro-0813",
+    name: "DeepSeek V4 Pro 0813",
+    configId: "255de820-3615-4921-a5fe-85b4af9e37a4",
+    modelId: "accounts/fireworks/models/deepseek-v4-pro-0813",
+    supportsVision: true,
+  },
+  {
+    id: "deepseek-v4-flash-0731",
+    name: "DeepSeek V4 Flash 0731",
+    configId: "22d8bd9a-3c48-4e27-a4f8-dc67ad2242b7",
+    modelId: "accounts/fireworks/models/deepseek-v4-flash-0731",
+    supportsVision: true,
  },
  {
    id: "gemini-3.1-pro-preview",
    name: "Gemini 3.1 Pro Preview",
-    configId: "d2bda5cd-881b-4044-aeb9-02a83cc0ca27",
+    configId: "17703a97-41a4-469d-b5d4-7356f1c28948",
    modelId: "google/gemini-3.1-pro-preview",
+    supportsVision: true,
  },
  {
-    id: "gemini-3.5-flash",
-    name: "Gemini 3.5 Flash",
-    configId: "c3a25aa0-ca48-4577-b52d-71282aacb687",
-    modelId: "google/gemini-3.5-flash",
-  },
-  {
-    id: "glm-5.2",
-    name: "GLM 5.2",
-    configId: "64a1fa3d-bf2e-4bb9-8c2b-fa76c218d636",
-    modelId: "accounts/fireworks/models/glm-5p2",
-  },
-  {
-    id: "gpt-5.5",
-    name: "GPT 5.5",
-    configId: "1762fbce-d5bf-4bf4-ba3d-8b1201f8e204",
-    modelId: "gpt-5.5",
-  },
-  {
-    id: "gpt-5.6-luna",
-    name: "GPT-5.6 Luna",
-    configId: "a9c45ba7-87fa-49a1-8165-76b0864c3a55",
-    modelId: "gpt-5.6-luna",
+    id: "gemini-3.7-flash",
+    name: "Gemini 3.7 Flash",
+    configId: "60754535-a5e8-4ae7-acf2-43d046771700",
+    modelId: "google/gemini-3.7-flash",
+    supportsVision: true,
  },
  {
    id: "gpt-5.6-sol",
    name: "GPT-5.6 Sol",
-    configId: "34c80712-def3-4db3-9e7a-f57b0324b43d",
+    configId: "4914e63d-ea29-45dc-9a85-c367b1ad0be5",
    modelId: "gpt-5.6-sol",
+    supportsVision: true,
  },
  {
    id: "gpt-5.6-terra",
    name: "GPT-5.6 Terra",
-    configId: "04f1a08c-42b2-4371-b6d8-75c50b9bb990",
+    configId: "4e627eb9-a199-4a90-8050-b734b6ee5fda",
    modelId: "gpt-5.6-terra",
+    supportsVision: true,
  },
  {
-    id: "xai-grok-4-5",
-    name: "Grok 4.5",
-    configId: "068b2ef2-e432-422b-98e5-5863a1852c47",
-    modelId: "grok-4.5",
+    id: "gpt-5.6-luna",
+    name: "GPT-5.6 Luna",
+    configId: "5eac2efb-7951-4da6-9a32-bfb31b1a7788",
+    modelId: "gpt-5.6-luna",
+    supportsVision: true,
  },
  {
-    id: "kimi-k2.6",
-    name: "Kimi K2.6",
-    configId: "placeholder-kimi-k2.6",
-    modelId: "accounts/fireworks/models/kimi-k2p6",
+    id: "xai-grok-4-6",
+    name: "Grok 4.6",
+    configId: "673e97ba-7b15-4213-8984-9b2477ee3409",
+    modelId: "grok-4.6",
+    supportsVision: true,
  },
  {
-    id: "kimi-k2.7-code",
-    name: "Kimi K2.7 Code",
-    configId: "placeholder-kimi-k2.7-code",
-    modelId: "accounts/fireworks/models/kimi-k2p7-code",
+    id: "kimi-k3",
+    name: "Kimi K3",
+    configId: "2a751e62-e281-4ab0-9be0-b05a6f8603db",
+    modelId: "accounts/fireworks/models/kimi-k3",
+    supportsVision: true,
+  },
+  {
+    id: "glm-5.2",
+    name: "GLM 5.2",
+    configId: "d2694d4d-4285-4d3c-ada5-aea5956375d4",
+    modelId: "accounts/fireworks/models/glm-5p2",
+    supportsVision: false,
  },
  {
    id: "minimax-m3",
    name: "Minimax M3",
-    configId: "placeholder-minimax-m3",
+    configId: "c4028069-0eb9-4a31-825c-a1cff9e5a085",
    modelId: "accounts/fireworks/models/minimax-m3",
-    supportsVision: true,
+    supportsVision: false,
  },
 ];

--- a/src/app/(dashboard)/dashboard/combos/page.tsx
+++ b/src/app/(dashboard)/dashboard/combos/page.tsx
@@ -2686,7 +2686,7 @@ function ComboFormModal({ isOpen, combo, onClose, onSave, activeProviders, combo
  };

  const FREE_STACK_PRESET_MODELS = [
-    { model: "agy/gemini-3.5-flash-low", weight: 0 },
+    { model: "agy/gemini-3.7-flash-low", weight: 0 },
    { model: "kr/claude-sonnet-4.5", weight: 0 },
    { model: "if/kimi-k2-thinking", weight: 0 },
    { model: "if/qwen3-coder-plus", weight: 0 },
--- a/src/app/api/providers/[id]/models/route.ts
+++ b/src/app/api/providers/[id]/models/route.ts
@@ -1768,8 +1768,8 @@ export async function GET(

      // Vertex AI lists models from the Generative Language `v1beta/models` endpoint, which both
      // Express-mode API keys (via ?key=) and Service Account JSON (via a minted OAuth Bearer
-      // token) can reach. This surfaces the full live catalog — including image models
-      // (imagen-*, gemini-*-image) absent from the static registry list.
+      // token) can reach. This surfaces the live catalog, including gemini-*-image models
+      // absent from the static registry list.
      const credential = (apiKey || "").trim();
      let queryKey: string | null = null;
      let bearerToken: string | null = null;
--- a/src/lib/providerModels/geminiModelsParser.ts
+++ b/src/lib/providerModels/geminiModelsParser.ts
@@ -3,15 +3,11 @@
 *
 * Each model's `supportedGenerationMethods` is mapped to OmniRoute endpoints:
 *   - generateContent / generateAnswer → "chat"
- *   - predict                          → "images"  (Imagen image generation)
 *   - predictLongRunning               → "video"   (Veo video generation)
 *   - embedContent                     → "embeddings"
 *   - bidiGenerateContent              → "audio"   (Live real-time audio)
 *
- * Model-id heuristics refine the long-running bucket because Google exposes both
- * Imagen and Veo via long-running methods on the same endpoint:
- *   - id contains "veo"    → ensure "video"
- *   - id contains "imagen" → force "images" (never "video")
+ * Model-id heuristics ensure Veo models remain in the video bucket.
 *
 * Note: `gemini-*-image` models (e.g. gemini-3-pro-image) generate images via the
 * regular `generateContent` path, so they stay "chat" (image output is a chat
@@ -25,7 +21,6 @@
 const METHOD_TO_ENDPOINT: Record<string, string> = {
  generateContent: "chat",
  embedContent: "embeddings",
-  predict: "images",
  predictLongRunning: "video",
  bidiGenerateContent: "audio",
  generateAnswer: "chat",
@@ -65,15 +60,10 @@ export function parseGeminiModelsList(data: any): GeminiDiscoveryModel[] {
    const id = ((m.name as string) || (m.id as string) || "").replace(/^models\//, "");
    const lowerId = id.toLowerCase();

-    // Google exposes Imagen (image) and Veo (video) via long-running methods; the
-    // method alone can't always distinguish them, so refine by model id.
+    // Keep Veo models in the video bucket even when the method list is incomplete.
    if (lowerId.includes("veo")) {
      endpoints.add("video");
    }
-    if (lowerId.includes("imagen")) {
-      endpoints.delete("video");
-      endpoints.add("images");
-    }

    if (endpoints.size === 0) endpoints.add("chat");

--- a/src/shared/constants/cliTools.ts
+++ b/src/shared/constants/cliTools.ts
@@ -100,7 +100,10 @@ export const CLI_TOOLS: Record<string, CliCatalogEntry> = {
    baseUrlSupport: "none",
    defaultCommand: "zcode",
    notes: [
-      { type: "info", text: "Uses the local ZCode app-server and its existing builtin:zai-coding-plan login." },
+      {
+        type: "info",
+        text: "Uses the local ZCode app-server and its existing builtin:zai-coding-plan login.",
+      },
      { type: "warning", text: "The response is buffered until the ZCode turn completes." },
    ],
  },
@@ -239,29 +242,21 @@ export const CLI_TOOLS: Record<string, CliCatalogEntry> = {
    acpSpawnable: false,
    baseUrlSupport: "none",
    modelAliases: [
-      "gemini-3.7-flash",
-      "gemini-3.6-flash-high",
-      "gemini-3.6-flash-medium",
-      "gemini-3.6-flash-low",
+      "gemini-3.7-flash-high",
+      "gemini-3.7-flash-medium",
+      "gemini-3.7-flash-low",
      "claude-opus-4-6-thinking",
      "claude-sonnet-4-6",
      "gemini-pro-agent",
      "gemini-3.1-pro-low",
-      "gemini-3-flash-agent",
-      "gemini-3.5-flash-low",
-      "gemini-3.5-flash-extra-low",
      "gpt-oss-120b-medium",
    ],
    defaultModels: [
-      createCliModel("gemini-3.7-flash", "Gemini 3.7 Flash"),
-      createCliModel("gemini-3.6-flash-high", "Gemini 3.6 Flash High"),
-      createCliModel("gemini-3.6-flash-medium", "Gemini 3.6 Flash Medium"),
-      createCliModel("gemini-3.6-flash-low", "Gemini 3.6 Flash Low"),
+      createCliModel("gemini-3.7-flash-high", "Gemini 3.7 Flash High"),
+      createCliModel("gemini-3.7-flash-medium", "Gemini 3.7 Flash Medium"),
+      createCliModel("gemini-3.7-flash-low", "Gemini 3.7 Flash Low"),
      createCliModel("gemini-pro-agent", "Gemini 3.1 Pro High"),
      createCliModel("gemini-3.1-pro-low", "Gemini 3.1 Pro Low"),
-      createCliModel("gemini-3-flash-agent", "Gemini 3.5 Flash High"),
-      createCliModel("gemini-3.5-flash-low", "Gemini 3.5 Flash Medium"),
-      createCliModel("gemini-3.5-flash-extra-low", "Gemini 3.5 Flash Low"),
      createCliModel("claude-sonnet-4-6", "Claude Sonnet 4.6"),
      createCliModel("claude-opus-4-6-thinking", "Claude Opus 4.6 Thinking"),
      createCliModel("gpt-oss-120b-medium", "GPT OSS 120B Medium"),
--- a/src/shared/constants/modelSpecs.ts
+++ b/src/shared/constants/modelSpecs.ts
@@ -153,7 +153,7 @@ export const MODEL_SPECS: Record<string, ModelSpec> = {
    aliases: ["openai/gpt-4o"],
  },

-  // ── Gemini 2.5 and 3.5 Flash series ──────────────────────────────
+  // ── Gemini 2.5 and provider-neutral 3.5 Flash series ─────────────
  "gemini-2.5-flash": {
    maxOutputTokens: 65536,
    contextWindow: 1048576,
@@ -174,24 +174,17 @@ export const MODEL_SPECS: Record<string, ModelSpec> = {
    thinkingBudgetCap: 0,
  },

-  // ── Gemini 3.7 Flash (Antigravity) — collapsed live id ──────────
-  // Upstream (fetchAvailableModels on daily-cloudcode-pa) also serves this model as a
-  // single `gemini-3.7-flash-tiered` id via the `gemini-3.7-flash` alias in
-  // antigravityModelAliases.ts. Registered independently of the suffixed tier ids below
-  // (#3696 uniqueness invariant: each public id resolves to a distinct upstream id).
-  "gemini-3.7-flash": {
-    ...GEMINI_35_FLASH_MODEL_SPEC,
-    defaultThinkingBudget: 8192,
-    thinkingBudgetCap: 65536,
-    supportsThinking: true,
-  },
-
-  // ── Gemini 3.7 / 3.6 Flash (Antigravity live tiers) ─────────────
+  // ── Gemini 3.7 Flash (current Antigravity/AGY live tiers) ─────────
  // The model id itself selects the upstream 10k/4k/1k reasoning tier. Antigravity
  // still rejects client-supplied thinking parameters, so keep the explicit-parameter
-  // capability aligned with the existing Gemini 3.5 tier ids.
+  // capability aligned with the existing Gemini Flash tier ids.
  "gemini-3.7-flash-high": { ...GEMINI_35_FLASH_MODEL_SPEC },
  "gemini-3.7-flash-medium": { ...GEMINI_35_FLASH_MODEL_SPEC },
+  "gemini-3.7-flash-low": { ...GEMINI_35_FLASH_MODEL_SPEC },
+
+  // Provider-neutral compatibility for providers that still serve Gemini 3.6.
+  // Antigravity/AGY availability is governed by their own provider catalogs and
+  // retirement filters; these shared specs must not be treated as an allowlist.
  "gemini-3.6-flash-high": { ...GEMINI_35_FLASH_MODEL_SPEC },
  "gemini-3.6-flash-medium": { ...GEMINI_35_FLASH_MODEL_SPEC },
  "gemini-3.6-flash-low": { ...GEMINI_35_FLASH_MODEL_SPEC },
--- a/src/shared/constants/pricing/frontier-labs.ts
+++ b/src/shared/constants/pricing/frontier-labs.ts
@@ -3,6 +3,7 @@
 * Pure data; merged by default-pricing.ts via spread (god-file decomposition; semantic split).
 */
 import {
+  GEMINI_3_7_FLASH_PROMO_PRICING,
  GPT_5_5_PRICING,
  GPT_5_6_LUNA_PRICING,
  GPT_5_6_SOL_PRICING,
@@ -226,6 +227,7 @@ export const DEFAULT_PRICING_FRONTIER = {
    "claude-opus-4": CLAUDE_OPUS_4_PRICING,
  },
  gemini: {
+    "gemini-3.7-flash": GEMINI_3_7_FLASH_PROMO_PRICING,
    // Gemini 3.1 Pro — novo flagship Google (2026-03-17)
    // Context: 1.050.000 tokens | Max Output: 65.536
    "gemini-3.1-pro": {
--- a/src/shared/constants/pricing/inference-hosts.ts
+++ b/src/shared/constants/pricing/inference-hosts.ts
@@ -140,12 +140,13 @@ export const DEFAULT_PRICING_INFERENCE = {
      reasoning: 1.75,
      cache_creation: 0.0583,
    },
-    "gemini-3-5-flash": {
-      input: 1.05,
-      output: 6.3,
-      cached: 0.1085,
-      reasoning: 6.3,
-      cache_creation: 0.0602,
+    // Live public catalog rate checked 2026-08-16.
+    "gemini-3.7-flash": {
+      input: 0.525,
+      output: 2.625,
+      cached: 0.13125,
+      reasoning: 2.625,
+      cache_creation: 0.017708,
    },
    "gemini-3-flash-preview": {
      input: 0.35,
@@ -218,7 +219,13 @@ export const DEFAULT_PRICING_INFERENCE = {
      reasoning: 21.0,
      cache_creation: 3.5,
    },
-    "gpt-5.6-terra": { input: 1.0, output: 6.0, cached: 0.125, reasoning: 6.0, cache_creation: 1.0 },
+    "gpt-5.6-terra": {
+      input: 1.0,
+      output: 6.0,
+      cached: 0.125,
+      reasoning: 6.0,
+      cache_creation: 1.0,
+    },
    "grok-4.5": { input: 1.4, output: 4.2, cached: 0.35, reasoning: 4.2, cache_creation: 1.4 },
    "kimi-k3": { input: 2.1, output: 10.5, cached: 0.2625, reasoning: 10.5, cache_creation: 2.1 },
    "minimax-m2.7": {
@@ -229,6 +236,15 @@ export const DEFAULT_PRICING_INFERENCE = {
      cache_creation: 0.168,
    },
  },
+  orcarouter: {
+    "google/gemini-3.6-flash": {
+      input: 1.5,
+      output: 7.5,
+      cached: 0.15,
+      reasoning: 7.5,
+      cache_creation: 1.5,
+    },
+  },
  synthetic: {
    "hf:openai/gpt-oss-120b": {
      input: 0.1,
--- a/src/shared/constants/pricing/oauth-subscriptions.ts
+++ b/src/shared/constants/pricing/oauth-subscriptions.ts
@@ -4,6 +4,7 @@
 */
 import {
  CLAUDE_OPUS_5_PRICING,
+  GEMINI_3_7_FLASH_PROMO_PRICING,
  GPT_5_3_CODEX_PRICING,
  GPT_5_5_PRICING,
  GPT_5_6_LUNA_PRICING,
@@ -11,6 +12,12 @@ import {
  GPT_5_6_TERRA_PRICING,
 } from "./shared-tiers";

+const ANTIGRAVITY_GEMINI_3_7_PRICING = {
+  "gemini-3.7-flash-low": GEMINI_3_7_FLASH_PROMO_PRICING,
+  "gemini-3.7-flash-medium": GEMINI_3_7_FLASH_PROMO_PRICING,
+  "gemini-3.7-flash-high": GEMINI_3_7_FLASH_PROMO_PRICING,
+};
+
 export const DEFAULT_PRICING_OAUTH = {
  cc: {
    "claude-fable-5": {
@@ -277,35 +284,6 @@ export const DEFAULT_PRICING_OAUTH = {
      reasoning: 18.0,
      cache_creation: 2.0,
    },
-    // Antigravity 2.0.4+ exposes Gemini 3.5 Flash as three public client ids
-    // (see ANTIGRAVITY_PUBLIC_MODELS in open-sse/config/antigravityModelAliases.ts):
-    //   gemini-3.5-flash-extra-low → "Gemini 3.5 Flash (Low)"
-    //   gemini-3-flash-agent   → "Gemini 3.5 Flash (High)"
-    //   gemini-3.5-flash-low   → "Gemini 3.5 Flash (Medium)"
-    // Without these rows, getPricingForModel("ag", id) returns null and downstream
-    // cost and quota calculations silently fall back to $0.
-    "gemini-3.5-flash-extra-low": {
-      input: 0.5,
-      output: 3.0,
-      cached: 0.03,
-      reasoning: 4.5,
-      cache_creation: 0.5,
-    },
-    "gemini-3-flash-agent": {
-      input: 0.5,
-      output: 3.0,
-      cached: 0.03,
-      reasoning: 4.5,
-      cache_creation: 0.5,
-    },
-    "gemini-3.5-flash-low": {
-      input: 0.5,
-      output: 3.0,
-      cached: 0.03,
-      reasoning: 4.5,
-      cache_creation: 0.5,
-    },
-    // `gemini-pro-agent` is the callable Antigravity id for Gemini 3.1 Pro (High).
    "gemini-pro-agent": {
      input: 4.0,
      output: 18.0,
@@ -313,40 +291,7 @@ export const DEFAULT_PRICING_OAUTH = {
      reasoning: 27.0,
      cache_creation: 4.0,
    },
-    // Gemini 3.6 Flash (released 2026-07-21) - three tier variants like 3.5 Flash
-    // (see ANTIGRAVITY_PUBLIC_MODELS / MODEL_SPECS which already carry the catalog
-    // entries). Without these rows, getPricingForModel("ag", id) returns null and
-    // downstream cost and quota calculations silently fall back to $0.
-    // Pricing: $1.50 input / $7.50 output / $0.15 cached per MTok. Thinking tokens
-    // billed at output rate.
-    "gemini-3.7-flash": {
-      input: 1.5,
-      output: 7.5,
-      cached: 0.15,
-      reasoning: 7.5,
-      cache_creation: 1.5,
-    },
-    "gemini-3.6-flash-low": {
-      input: 1.5,
-      output: 7.5,
-      cached: 0.15,
-      reasoning: 7.5,
-      cache_creation: 1.5,
-    },
-    "gemini-3.6-flash-medium": {
-      input: 1.5,
-      output: 7.5,
-      cached: 0.15,
-      reasoning: 7.5,
-      cache_creation: 1.5,
-    },
-    "gemini-3.6-flash-high": {
-      input: 1.5,
-      output: 7.5,
-      cached: 0.15,
-      reasoning: 7.5,
-      cache_creation: 1.5,
-    },
+    ...ANTIGRAVITY_GEMINI_3_7_PRICING,
    "claude-sonnet-4-6": {
      input: 3.0,
      output: 15.0,
@@ -369,6 +314,8 @@ export const DEFAULT_PRICING_OAUTH = {
      cache_creation: 0.5,
    },
  },
+  antigravity: ANTIGRAVITY_GEMINI_3_7_PRICING,
+  agy: ANTIGRAVITY_GEMINI_3_7_PRICING,
  gh: {
    "claude-opus-5": CLAUDE_OPUS_5_PRICING,
    "gpt-5": {
@@ -441,6 +388,7 @@ export const DEFAULT_PRICING_OAUTH = {
      reasoning: 4.5,
      cache_creation: 0.5,
    },
+    "gemini-3.7-flash": GEMINI_3_7_FLASH_PROMO_PRICING,
    "gemini-2.5-pro": {
      input: 2.0,
      output: 12.0,
--- a/src/shared/constants/pricing/shared-tiers.ts
+++ b/src/shared/constants/pricing/shared-tiers.ts
@@ -41,6 +41,17 @@ export const GPT_5_6_LUNA_PRICING = {
  cache_creation: 1.25,
 };

+// Gemini 3.7 Flash introductory pricing through 2026-12-31. Google and
+// GitHub Copilot both publish the same current token rates; standard pricing
+// becomes $1.50/$7.50/$0.15 (input/output/cached input) on 2027-01-01.
+export const GEMINI_3_7_FLASH_PROMO_PRICING = {
+  input: 0.75,
+  output: 3.75,
+  cached: 0.075,
+  reasoning: 3.75,
+  cache_creation: 0.75,
+};
+
 export const CLAUDE_FABLE_5_PRICING = {
  input: 15.0,
  output: 75.0,
--- a/tests/unit/account-fallback-service.test.ts
+++ b/tests/unit/account-fallback-service.test.ts
@@ -127,7 +127,7 @@ test("checkFallbackError locks Antigravity quota-reached 429 for the full reset
    429,
    message,
    0,
-    "gemini-3-flash-agent",
+    "gemini-3.7-flash-high",
    "antigravity",
    null,
    makeProfile({ useUpstreamRetryHints: true })
@@ -143,7 +143,7 @@ test("checkFallbackError locks Antigravity quota-reached 429 for the full reset
 test("recordModelLockoutFailure honors a multi-day exactCooldownMs (under 30-day cap)", () => {
  const provider = "antigravity";
  const connectionId = "conn-quota-window";
-  const model = "gemini-3-flash-agent";
+  const model = "gemini-3.7-flash-high";
  const exactCooldownMs = (164 * 3600 + 27 * 60 + 24) * 1000;

  clearModelLock(provider, connectionId, model);
--- a/tests/unit/agy-provider.test.ts
+++ b/tests/unit/agy-provider.test.ts
@@ -51,12 +51,15 @@ test("agy ships its own live callable model catalog", () => {
  const ids = REGISTRY.agy.models.map((m) => m.id);
  assert.ok(ids.includes("claude-opus-4-6-thinking"), "must expose Claude Opus 4.6 Thinking");
  assert.ok(ids.includes("claude-sonnet-4-6"), "must expose Claude Sonnet 4.6");
-  assert.ok(ids.includes("gemini-3.6-flash-low"), "must expose Gemini 3.6 Flash Low");
-  assert.ok(ids.includes("gemini-3.6-flash-medium"), "must expose Gemini 3.6 Flash Medium");
-  assert.ok(ids.includes("gemini-3.6-flash-high"), "must expose Gemini 3.6 Flash High");
-  assert.ok(ids.includes("gemini-3.5-flash-extra-low"), "must expose upstream Flash Low tier");
-  assert.ok(ids.includes("gemini-3.5-flash-low"), "must expose upstream Flash Medium tier");
-  assert.ok(ids.includes("gemini-3-flash-agent"), "must expose upstream Flash High tier");
+  assert.ok(ids.includes("gemini-3.7-flash-low"), "must expose Gemini 3.7 Flash Low");
+  assert.ok(ids.includes("gemini-3.7-flash-medium"), "must expose Gemini 3.7 Flash Medium");
+  assert.ok(ids.includes("gemini-3.7-flash-high"), "must expose Gemini 3.7 Flash High");
+  assert.ok(!ids.includes("gemini-3.6-flash-low"));
+  assert.ok(!ids.includes("gemini-3.6-flash-medium"));
+  assert.ok(!ids.includes("gemini-3.6-flash-high"));
+  assert.ok(!ids.includes("gemini-3.5-flash-extra-low"));
+  assert.ok(!ids.includes("gemini-3.5-flash-low"));
+  assert.ok(!ids.includes("gemini-3-flash-agent"));
  assert.ok(!ids.includes("gemini-3.5-flash-medium"));
  assert.ok(!ids.includes("gemini-3.5-flash-high"));
  assert.ok(!ids.includes("gemini-3.5-flash-preview"));
@@ -64,8 +67,8 @@ test("agy ships its own live callable model catalog", () => {
  assert.ok(!ids.includes("gemini-3.1-pro-high"), "must not expose rejected Pro High id");
  assert.ok(ids.includes("gemini-pro-agent"), "must expose callable Pro High id");
  assert.ok(!ids.includes("gemini-2.5-pro"), "must not expose unavailable Gemini 2.5 Pro");
-  assert.ok(ids.includes("gemini-2.5-flash"), "must retain working Gemini 2.5 Flash");
-  assert.ok(ids.includes("gemini-2.5-flash-lite"), "must retain working Gemini 2.5 Flash Lite");
+  assert.ok(!ids.includes("gemini-2.5-flash"));
+  assert.ok(!ids.includes("gemini-2.5-flash-lite"));
  // Tab-completion models are not chat-callable and must be excluded.
  assert.ok(!ids.includes("tab_flash_lite_preview"));
  assert.ok(!ids.includes("tab_jump_flash_lite_preview"));
@@ -75,15 +78,18 @@ test("agy ships its own live callable model catalog", () => {
 test("agy model helpers resolve catalog ids and display names", () => {
  assert.equal(isUserCallableAgyModelId("claude-opus-4-6-thinking"), true);
  assert.equal(isUserCallableAgyModelId("gemini-2.5-pro"), false);
-  assert.equal(isUserCallableAgyModelId("gemini-2.5-flash"), true);
+  assert.equal(isUserCallableAgyModelId("gemini-2.5-flash"), false);
  assert.equal(isUserCallableAgyModelId("gemini-3.1-pro-high"), false);
  assert.equal(isUserCallableAgyModelId("gemini-pro-agent"), true);
-  assert.equal(isUserCallableAgyModelId("gemini-3.6-flash-low"), true);
-  assert.equal(isUserCallableAgyModelId("gemini-3.6-flash-medium"), true);
-  assert.equal(isUserCallableAgyModelId("gemini-3.6-flash-high"), true);
-  assert.equal(isUserCallableAgyModelId("gemini-3.5-flash-extra-low"), true);
-  assert.equal(isUserCallableAgyModelId("gemini-3.5-flash-low"), true);
-  assert.equal(isUserCallableAgyModelId("gemini-3-flash-agent"), true);
+  assert.equal(isUserCallableAgyModelId("gemini-3.7-flash-low"), true);
+  assert.equal(isUserCallableAgyModelId("gemini-3.7-flash-medium"), true);
+  assert.equal(isUserCallableAgyModelId("gemini-3.7-flash-high"), true);
+  assert.equal(isUserCallableAgyModelId("gemini-3.6-flash-low"), false);
+  assert.equal(isUserCallableAgyModelId("gemini-3.6-flash-medium"), false);
+  assert.equal(isUserCallableAgyModelId("gemini-3.6-flash-high"), false);
+  assert.equal(isUserCallableAgyModelId("gemini-3.5-flash-extra-low"), false);
+  assert.equal(isUserCallableAgyModelId("gemini-3.5-flash-low"), false);
+  assert.equal(isUserCallableAgyModelId("gemini-3-flash-agent"), false);
  assert.equal(isUserCallableAgyModelId("gemini-3.5-flash-medium"), false);
  assert.equal(isUserCallableAgyModelId("gemini-3.5-flash-high"), false);
  assert.equal(isUserCallableAgyModelId("gemini-3.5-flash-preview"), false);
@@ -94,23 +100,20 @@ test("agy model helpers resolve catalog ids and display names", () => {
    "Claude Opus 4.6 (Thinking)"
  );
  assert.equal(getClientVisibleAgyModelName("gemini-pro-agent"), "Gemini 3.1 Pro (High)");
+  assert.equal(getClientVisibleAgyModelName("gemini-3.7-flash-low"), "Gemini 3.7 Flash (Low)");
  assert.equal(
-    getClientVisibleAgyModelName("gemini-3.5-flash-extra-low"),
-    "Gemini 3.5 Flash (Low)"
+    getClientVisibleAgyModelName("gemini-3.7-flash-medium"),
+    "Gemini 3.7 Flash (Medium)"
  );
-  assert.equal(getClientVisibleAgyModelName("gemini-3.6-flash-low"), "Gemini 3.6 Flash (Low)");
-  assert.equal(
-    getClientVisibleAgyModelName("gemini-3.6-flash-medium"),
-    "Gemini 3.6 Flash (Medium)"
-  );
-  assert.equal(getClientVisibleAgyModelName("gemini-3.6-flash-high"), "Gemini 3.6 Flash (High)");
-  assert.equal(getClientVisibleAgyModelName("gemini-3.5-flash-low"), "Gemini 3.5 Flash (Medium)");
-  assert.equal(getClientVisibleAgyModelName("gemini-3-flash-agent"), "Gemini 3.5 Flash (High)");
+  assert.equal(getClientVisibleAgyModelName("gemini-3.7-flash-high"), "Gemini 3.7 Flash (High)");
  assert.equal(getClientVisibleAgyModelName("unknown-model", "Fallback"), "Fallback");
 });

 test("agy live discovery accepts new chat models while excluding tab-completion models", () => {
  assert.equal(isDiscoverableAgyModelId("gemini-new-live-tier"), true);
+  assert.equal(isDiscoverableAgyModelId("gemini-3.6-flash-high"), false);
+  assert.equal(isDiscoverableAgyModelId("gemini-3-flash-agent"), false);
+  assert.equal(isDiscoverableAgyModelId("gemini-2.5-flash"), false);
  assert.equal(isDiscoverableAgyModelId("tab_flash_lite_preview"), false);
  assert.equal(isDiscoverableAgyModelId("tab_jump_flash_lite_preview"), false);
  assert.equal(isDiscoverableAgyModelId(""), false);
--- a/tests/unit/agy-usage-quota.test.ts
+++ b/tests/unit/agy-usage-quota.test.ts
@@ -18,7 +18,7 @@ test("getUsageForProvider routes agy through the Antigravity usage implementatio
    new Response(
      JSON.stringify({
        models: {
-          "gemini-3-flash-agent": {
+          "gemini-3.7-flash-high": {
            quotaInfo: {
              remainingFraction: 0.75,
              resetTime: "2026-06-06T00:00:00Z",
@@ -48,7 +48,7 @@ test("getUsageForProvider routes agy through the Antigravity usage implementatio
    );
    assert.ok("quotas" in result, "agy should return quota data when upstream responds");

-    const quota = (result as { quotas: Record<string, any> }).quotas["gemini-3-flash-agent"];
+    const quota = (result as { quotas: Record<string, any> }).quotas["gemini-3.7-flash-high"];
    assert.ok(quota, "should expose the upstream agy per-model quota");
    assert.equal(quota.remainingPercentage, 75);
    assert.equal(
@@ -65,7 +65,7 @@ test("parseQuotaData treats agy quota payloads like Antigravity", () => {
  const parsed = providerLimitUtils.parseQuotaData("agy", {
    quotas: {
      credits: { remaining: 42 },
-      "gemini-3-flash-agent": {
+      "gemini-3.7-flash-high": {
        used: 250,
        total: 1000,
        remainingPercentage: 75,
@@ -83,7 +83,7 @@ test("parseQuotaData treats agy quota payloads like Antigravity", () => {
  assert.ok(credits, "credits quota should be rendered");
  assert.equal(credits.isCredits, true);

-  const modelQuota = parsed.find((quota: any) => quota.name === "gemini-3-flash-agent");
+  const modelQuota = parsed.find((quota: any) => quota.name === "gemini-3.7-flash-high");
  assert.ok(modelQuota, "model quota should be rendered");
  assert.equal(modelQuota.remainingPercentage, 75);
 });
--- a/tests/unit/antigravity-local-usage-fallback-3821.test.ts
+++ b/tests/unit/antigravity-local-usage-fallback-3821.test.ts
@@ -7,7 +7,7 @@
 * the fallback queries and asserts the flip — the regression guard for the id contract.
 *
 * Contract note: the fallback queries `usage_history WHERE model = <public model id>`
- * (e.g. gemini-3-flash-agent), so the executor MUST log usage under that same model id
+ * (e.g. gemini-3.7-flash-high), so the executor MUST log usage under that same model id
 * for the fallback to fire. This test pins exactly that join.
 */
 import test from "node:test";
@@ -47,7 +47,7 @@ test("Antigravity fetchAvailableModels(used=0) → localUsageHistory when usage_
  db.prepare(
    `INSERT INTO usage_history (provider, model, connection_id, tokens_input, tokens_output, tokens_reasoning, success, timestamp)
     VALUES (?, ?, ?, ?, ?, ?, 1, ?)`
-  ).run("antigravity", "gemini-3-flash-agent", "conn-local-1", 1000, 1500, 500, seededTimestamp);
+  ).run("antigravity", "gemini-3.7-flash-high", "conn-local-1", 1000, 1500, 500, seededTimestamp);
  // Total seeded tokens = 3000 → ceil(3000/1000) = 3 units used.

  globalThis.fetch = (async (input: any) => {
@@ -61,7 +61,7 @@ test("Antigravity fetchAvailableModels(used=0) → localUsageHistory when usage_
      ok: true,
      json: async () => ({
        models: {
-          "gemini-3-flash-agent": {
+          "gemini-3.7-flash-high": {
            quotaInfo: { remainingFraction: 1.0, resetTime },
          },
        },
@@ -79,8 +79,8 @@ test("Antigravity fetchAvailableModels(used=0) → localUsageHistory when usage_

  const result = await getUsageForProvider(connection, { forceRefresh: true });
  assert.ok(result && "quotas" in result, "should return quotas");
-  const quota = (result as any).quotas["gemini-3-flash-agent"];
-  assert.ok(quota, "should have the gemini-3-flash-agent quota");
+  const quota = (result as any).quotas["gemini-3.7-flash-high"];
+  assert.ok(quota, "should have the gemini-3.7-flash-high quota");
  assert.equal(quota.quotaSource, "localUsageHistory", "stale full bucket replaced by local usage");
  assert.equal(quota.used, 3, "3000 seeded tokens → 3 units used");
 });
@@ -99,7 +99,7 @@ test("Antigravity stays fetchAvailableModels when usage_history has no matching
      ok: true,
      json: async () => ({
        models: {
-          "gemini-3-flash-agent": { quotaInfo: { remainingFraction: 1.0, resetTime } },
+          "gemini-3.7-flash-high": { quotaInfo: { remainingFraction: 1.0, resetTime } },
        },
      }),
    } as Response;
@@ -114,7 +114,7 @@ test("Antigravity stays fetchAvailableModels when usage_history has no matching
  };

  const result = await getUsageForProvider(connection, { forceRefresh: true });
-  const quota = (result as any).quotas["gemini-3-flash-agent"];
+  const quota = (result as any).quotas["gemini-3.7-flash-high"];
  assert.ok(quota, "should have the quota");
  assert.equal(quota.quotaSource, "fetchAvailableModels", "no local rows → keep the catalog view");
  assert.equal(quota.used, 0, "full bucket stays at 0 used");
--- a/tests/unit/antigravity-model-aliases.test.ts
+++ b/tests/unit/antigravity-model-aliases.test.ts
@@ -21,15 +21,16 @@ const EXPECTED_FLASH_TIERS = [
  ["gemini-3.7-flash", "Gemini 3.7 Flash"],
  ["gemini-3.7-flash-high", "Gemini 3.7 Flash (High)"],
  ["gemini-3.7-flash-medium", "Gemini 3.7 Flash (Medium)"],
-  ["gemini-3.6-flash-low", "Gemini 3.6 Flash (Low)"],
-  ["gemini-3.6-flash-medium", "Gemini 3.6 Flash (Medium)"],
-  ["gemini-3.6-flash-high", "Gemini 3.6 Flash (High)"],
-  ["gemini-3.5-flash-extra-low", "Gemini 3.5 Flash (Low)"],
-  ["gemini-3.5-flash-low", "Gemini 3.5 Flash (Medium)"],
-  ["gemini-3-flash-agent", "Gemini 3.5 Flash (High)"],
+  ["gemini-3.7-flash-low", "Gemini 3.7 Flash (Low)"],
 ] as const;

 const RETIRED_FLASH_IDS = [
+  "gemini-3.6-flash-low",
+  "gemini-3.6-flash-medium",
+  "gemini-3.6-flash-high",
+  "gemini-3.5-flash-extra-low",
+  "gemini-3.5-flash-low",
+  "gemini-3-flash-agent",
  "gemini-3.5-flash-medium",
  "gemini-3.5-flash-high",
  "gemini-3.5-flash-preview",
@@ -44,6 +45,9 @@ test("toClientAntigravityQuotaModelId preserves upstream Gemini Flash bucket IDs
  // Retired preview buckets are dropped (hidden from clients).
  assert.equal(toClientAntigravityQuotaModelId("gemini-3.5-flash-preview"), null);
  assert.equal(toClientAntigravityQuotaModelId("gemini-3-flash-preview"), null);
+  for (const retiredId of RETIRED_FLASH_IDS) {
+    assert.equal(toClientAntigravityQuotaModelId(retiredId), null);
+  }
  assert.equal(toClientAntigravityQuotaModelId(""), null);
 });

@@ -86,9 +90,9 @@ test("isUserCallableAntigravityModelId only allows public chat-capable model IDs
  }
  assert.equal(isUserCallableAntigravityModelId("gemini-3.1-flash-lite"), true);
  assert.equal(isUserCallableAntigravityModelId("gemini-2.5-pro"), false);
-  assert.equal(isUserCallableAntigravityModelId("gemini-2.5-flash"), true);
-  assert.equal(isUserCallableAntigravityModelId("gemini-2.5-flash-lite"), true);
-  assert.equal(isUserCallableAntigravityModelId("gemini-2.5-flash-thinking"), true);
+  assert.equal(isUserCallableAntigravityModelId("gemini-2.5-flash"), false);
+  assert.equal(isUserCallableAntigravityModelId("gemini-2.5-flash-lite"), false);
+  assert.equal(isUserCallableAntigravityModelId("gemini-2.5-flash-thinking"), false);
  assert.equal(isUserCallableAntigravityModelId("gemini-pro-agent"), true);
  // #3184: Claude IS user-callable through the Antigravity OAuth provider (same backend as
  // `agy`, verified empirically). An earlier assumption that it was removed in Antigravity
@@ -109,6 +113,10 @@ test("isDiscoverableAntigravityModelId accepts new live chat models without a st
  assert.equal(isDiscoverableAntigravityModelId("claude-sonnet-5"), true);
  assert.equal(isDiscoverableAntigravityModelId("gemini-new-live-tier"), true);

+  for (const retiredId of RETIRED_FLASH_IDS) {
+    assert.equal(isDiscoverableAntigravityModelId(retiredId), false);
+  }
+
  assert.equal(isDiscoverableAntigravityModelId("tab_flash_lite_preview"), false);
  assert.equal(isDiscoverableAntigravityModelId("gemini-3.1-flash-image"), false);
  assert.equal(isDiscoverableAntigravityModelId("gemini-3.1-flash-tts-preview"), false);
@@ -150,15 +158,6 @@ test("ANTIGRAVITY_PUBLIC_MODELS exposes current live names and capabilities", ()
  for (const retiredId of RETIRED_FLASH_IDS) {
    assert.equal(getPublicModel(retiredId), undefined);
  }
-  assert.equal(getClientVisibleAntigravityModelName("gemini-2.5-flash"), "Gemini 2.5 Flash");
-  assert.equal(
-    getClientVisibleAntigravityModelName("gemini-2.5-flash-lite"),
-    "Gemini 2.5 Flash Lite"
-  );
-  assert.equal(
-    getClientVisibleAntigravityModelName("gemini-2.5-flash-thinking"),
-    "Gemini 2.5 Flash Thinking"
-  );
  assert.deepEqual(getPublicModel("gpt-oss-120b-medium"), {
    id: "gpt-oss-120b-medium",
    name: "GPT-OSS 120B (Medium)",
--- a/tests/unit/antigravity-native-toolcall-collect.test.ts
+++ b/tests/unit/antigravity-native-toolcall-collect.test.ts
@@ -7,7 +7,7 @@
 * part (usually carrying a `thoughtSignature`), so any tools request collected
 * to an empty stream and was rewritten into a synthetic 502 "Provider returned
 * empty content" — breaking Chatwit Captain Copilot / reply suggestions on
- * agy/gemini-3.5-flash-low while plain text completions kept working.
+ * agy/gemini-3.7-flash-low while plain text completions kept working.
 */
 import test from "node:test";
 import assert from "node:assert/strict";
--- a/tests/unit/antigravity-per-model-output-cap.test.ts
+++ b/tests/unit/antigravity-per-model-output-cap.test.ts
@@ -82,6 +82,21 @@ test("an unknown model id falls back to the conservative ceiling", () => {
  assert.equal(clampFor("no-such-model-xyz", 65535), MAX_ANTIGRAVITY_OUTPUT_TOKENS);
 });

+test("retired Antigravity Flash ids cannot inherit provider-neutral output caps", () => {
+  for (const modelId of [
+    "gemini-3.6-flash-low",
+    "gemini-3.6-flash-medium",
+    "gemini-3.6-flash-high",
+    "gemini-3.5-flash-extra-low",
+    "gemini-3.5-flash-low",
+    "gemini-3.5-flash-medium",
+    "gemini-3.5-flash-high",
+    "gemini-3-flash-agent",
+  ]) {
+    assert.equal(clampFor(modelId, 65536), MAX_ANTIGRAVITY_OUTPUT_TOKENS, modelId);
+  }
+});
+
 test("a missing or empty model id falls back too", () => {
  assert.equal(clampFor(undefined, 65535), MAX_ANTIGRAVITY_OUTPUT_TOKENS);
  assert.equal(clampFor(null, 65535), MAX_ANTIGRAVITY_OUTPUT_TOKENS);
@@ -211,7 +226,7 @@ test("a provider-prefixed model id resolves to the model's ceiling, not the fall
  const cases: Array<[string, number]> = [
    ["agy/gemini-3.1-pro-high", 65535],
    ["antigravity/gemini-3.1-pro-high", 65535],
-    ["agy/gemini-3.6-flash-high", 65536],
+    ["agy/gemini-3.7-flash-high", 65536],
    ["agy/gpt-oss-120b-medium", 32768],
  ];

--- a/tests/unit/antigravity-quota-host-8965.test.ts
+++ b/tests/unit/antigravity-quota-host-8965.test.ts
@@ -89,7 +89,7 @@ test("#8965: quota reads use the runtime host (daily-cloudcode-pa), not cloudcod
          json: async () => ({
            buckets: [
              {
-                modelId: "gemini-3-flash-agent",
+                modelId: "gemini-3.7-flash-high",
                remainingFraction: 0.4,
                resetTime: RESET_IN_2_HOURS,
              },
@@ -103,10 +103,10 @@ test("#8965: quota reads use the runtime host (daily-cloudcode-pa), not cloudcod
          ok: true,
          json: async () => ({
            models: {
-              "gemini-3-flash-agent": {
+              "gemini-3.7-flash-high": {
                quotaInfo: { remainingFraction: 1.0, resetTime: RESET_IN_2_HOURS },
              },
-              "gemini-3.5-flash-low": {
+              "gemini-3.7-flash-medium": {
                quotaInfo: { remainingFraction: 0.8, resetTime: RESET_IN_2_HOURS },
              },
            },
@@ -148,9 +148,9 @@ test("#8965: quota reads use the runtime host (daily-cloudcode-pa), not cloudcod

  // The per-model quota should come from retrieveUserQuota (the live source),
  // NOT fetchAvailableModels (the stale catalog fallback).
-  assert.ok(quotas["gemini-3-flash-agent"], "gemini-3-flash-agent quota present");
+  assert.ok(quotas["gemini-3.7-flash-high"], "gemini-3.7-flash-high quota present");
  assert.equal(
-    quotas["gemini-3-flash-agent"].quotaSource,
+    quotas["gemini-3.7-flash-high"].quotaSource,
    "retrieveUserQuota",
    "quota source is retrieveUserQuota (live), not fetchAvailableModels"
  );
@@ -197,7 +197,7 @@ test("#8965 behavioral impact: live quota source + weekly bucket unreachable whe
          json: async () => ({
            buckets: [
              {
-                modelId: "gemini-3-flash-agent",
+                modelId: "gemini-3.7-flash-high",
                remainingFraction: 0.4,
                resetTime: RESET_IN_2_HOURS,
              },
@@ -211,7 +211,7 @@ test("#8965 behavioral impact: live quota source + weekly bucket unreachable whe
          ok: true,
          json: async () => ({
            models: {
-              "gemini-3-flash-agent": {
+              "gemini-3.7-flash-high": {
                quotaInfo: { remainingFraction: 1.0, resetTime: RESET_IN_2_HOURS },
              },
            },
@@ -250,9 +250,9 @@ test("#8965 behavioral impact: live quota source + weekly bucket unreachable whe
  const quotas = (result as UsageResult).quotas;

  // The per-model quota MUST come from retrieveUserQuota — the live signal.
-  assert.ok(quotas["gemini-3-flash-agent"], "gemini-3-flash-agent quota present");
+  assert.ok(quotas["gemini-3.7-flash-high"], "gemini-3.7-flash-high quota present");
  assert.equal(
-    quotas["gemini-3-flash-agent"].quotaSource,
+    quotas["gemini-3.7-flash-high"].quotaSource,
    "retrieveUserQuota",
    "quota source is retrieveUserQuota (live), not fetchAvailableModels"
  );
@@ -260,4 +260,4 @@ test("#8965 behavioral impact: live quota source + weekly bucket unreachable whe
  // The weekly group quota MUST also be present because retrieveUserQuotaSummary
  // was served by the runtime host.
  assert.ok(quotas.gemini_weekly, "weekly group quota present");
-});
+});
--- a/tests/unit/antigravity-quota-skipping.test.ts
+++ b/tests/unit/antigravity-quota-skipping.test.ts
@@ -21,7 +21,7 @@ test("isQuotaExhaustedForRequest isolates Claude and Gemini quota families for a
  // Simulate Claude Opus being exhausted, while Gemini is NOT.
  quotaCache.setQuotaCache(connectionId, "antigravity", {
    "claude-opus-4-6-thinking": { remainingPercentage: 0, resetAt: null },
-    "gemini-3.5-flash-high": { remainingPercentage: 100, resetAt: null },
+    "gemini-3.7-flash-high": { remainingPercentage: 100, resetAt: null },
  });

  // Verify that Claude models are considered exhausted.
@@ -49,7 +49,7 @@ test("isQuotaExhaustedForRequest isolates Claude and Gemini quota families for a
    quotaCache.isQuotaExhaustedForRequest(
      connectionId,
      "antigravity",
-      "antigravity/gemini-3.5-flash-high"
+      "antigravity/gemini-3.7-flash-high"
    ),
    false,
    "Gemini Flash should NOT be exhausted"
@@ -58,7 +58,7 @@ test("isQuotaExhaustedForRequest isolates Claude and Gemini quota families for a
    quotaCache.isQuotaExhaustedForRequest(
      connectionId,
      "antigravity",
-      "antigravity/gemini-2.5-pro"
+      "antigravity/gemini-pro-agent"
    ),
    false,
    "Gemini Pro should share Gemini family quota and NOT be exhausted"
@@ -68,7 +68,7 @@ test("isQuotaExhaustedForRequest isolates Claude and Gemini quota families for a
  const connectionIdAgy = "conn-agy-test";
  quotaCache.setQuotaCache(connectionIdAgy, "agy", {
    "claude-opus-4-6-thinking": { remainingPercentage: 0, resetAt: null },
-    "gemini-3.5-flash-high": { remainingPercentage: 100, resetAt: null },
+    "gemini-3.7-flash-high": { remainingPercentage: 100, resetAt: null },
  });

  assert.equal(
@@ -77,7 +77,7 @@ test("isQuotaExhaustedForRequest isolates Claude and Gemini quota families for a
    "Claude Opus under 'agy' should be exhausted"
  );
  assert.equal(
-    quotaCache.isQuotaExhaustedForRequest(connectionIdAgy, "agy", "agy/gemini-3.5-flash-high"),
+    quotaCache.isQuotaExhaustedForRequest(connectionIdAgy, "agy", "agy/gemini-3.7-flash-high"),
    false,
    "Gemini Flash under 'agy' should NOT be exhausted"
  );
@@ -112,8 +112,8 @@ test("isQuotaExhaustedForRequest isolates Claude and Gemini quota families for a
 test("isQuotaExhaustedForRequest scopes gemini exhaustion to the requested model, not sibling models", () => {
  const connectionId = "conn-gemini-sibling-test";
  quotaCache.setQuotaCache(connectionId, "antigravity", {
-    "gemini-3.6-flash-medium": { remainingPercentage: 0, resetAt: null },
-    "gemini-2.5-pro": { remainingPercentage: 100, resetAt: null },
+    "gemini-3.7-flash-medium": { remainingPercentage: 0, resetAt: null },
+    "gemini-pro-agent": { remainingPercentage: 100, resetAt: null },
    gemini_weekly: { remainingPercentage: 0, resetAt: null },
  });

@@ -121,33 +121,33 @@ test("isQuotaExhaustedForRequest scopes gemini exhaustion to the requested model
    quotaCache.isQuotaExhaustedForRequest(
      connectionId,
      "antigravity",
-      "antigravity/gemini-3.6-flash-medium"
+      "antigravity/gemini-3.7-flash-medium"
    ),
    true,
-    "gemini-3.6 at 0% should be exhausted even when gemini-2.5-pro still has quota"
+    "gemini-3.7 at 0% should be exhausted even when gemini-pro-agent still has quota"
  );
  assert.equal(
    quotaCache.isQuotaExhaustedForRequest(
      connectionId,
      "antigravity",
-      "antigravity/gemini-2.5-pro"
+      "antigravity/gemini-pro-agent"
    ),
    false,
-    "gemini-2.5-pro should remain available when only gemini-3.6 is depleted"
+    "gemini-pro-agent should remain available when only gemini-3.7 Flash is depleted"
  );
 });

 test("isQuotaExhaustedForRequest treats near-zero remaining as exhausted at default threshold", () => {
  const connectionId = "conn-near-zero-test";
  quotaCache.setQuotaCache(connectionId, "antigravity", {
-    "gemini-3.6-flash-medium": { remainingPercentage: 0.00000167, resetAt: null },
+    "gemini-3.7-flash-medium": { remainingPercentage: 0.00000167, resetAt: null },
  });

  assert.equal(
    quotaCache.isQuotaExhaustedForRequest(
      connectionId,
      "antigravity",
-      "antigravity/gemini-3.6-flash-medium"
+      "antigravity/gemini-3.7-flash-medium"
    ),
    true,
    "effectively-zero remaining should count as exhausted"
--- a/tests/unit/antigravity-retired-public-models.test.ts
+++ b/tests/unit/antigravity-retired-public-models.test.ts
@@ -17,44 +17,40 @@ import { CLI_TOOLS } from "../../src/shared/constants/cliTools.ts";

 const RETIRED_PUBLIC_MODELS = [
  "gemini-3-pro-preview",
+  "gemini-3.6-flash-high",
+  "gemini-3.6-flash-medium",
+  "gemini-3.6-flash-low",
+  "gemini-3-flash-agent",
+  "gemini-3.5-flash-low",
+  "gemini-3.5-flash-extra-low",
  "gemini-2.5-pro",
+  "gemini-2.5-flash-thinking",
+  "gemini-2.5-flash",
+  "gemini-2.5-flash-lite",
  "gemini-2.5-computer-use-preview-10-2025",
 ] as const;

 const EXPECTED_LEADING_MODEL_ORDER = [
-  "gemini-3.6-flash-high",
-  "gemini-3.6-flash-medium",
-  "gemini-3.6-flash-low",
-  "claude-opus-4-6-thinking",
-  "claude-sonnet-4-6",
-  "gemini-pro-agent",
-  "gemini-3.1-pro-low",
-  "gemini-3-flash-agent",
-  "gemini-3.5-flash-low",
-  "gemini-3.5-flash-extra-low",
-] as const;
-
-const EXPECTED_ANTIGRAVITY_LEADING_MODEL_ORDER = [
  "gemini-3.7-flash-high",
  "gemini-3.7-flash-medium",
-  ...EXPECTED_LEADING_MODEL_ORDER,
+  "gemini-3.7-flash-low",
+  "gemini-pro-agent",
+  "gemini-3.1-pro-low",
+  "gemini-3.1-flash-lite",
+  "claude-opus-4-6-thinking",
+  "claude-sonnet-4-6",
+  "gpt-oss-120b-medium",
 ] as const;

 const ACTIVE_FLASH_MODEL_IDS = [
-  "gemini-3-flash-agent",
-  "gemini-3.5-flash-low",
-  "gemini-3.5-flash-extra-low",
-] as const;
-
-const CURRENT_36_FLASH_MODEL_IDS = [
-  "gemini-3.6-flash-high",
-  "gemini-3.6-flash-medium",
-  "gemini-3.6-flash-low",
+  "gemini-3.7-flash-high",
+  "gemini-3.7-flash-medium",
+  "gemini-3.7-flash-low",
 ] as const;

 test("Antigravity and AGY place their live Gemini Flash tiers first", () => {
  for (const [provider, models, expectedOrder] of [
-    ["antigravity", ANTIGRAVITY_PUBLIC_MODELS, EXPECTED_ANTIGRAVITY_LEADING_MODEL_ORDER],
+    ["antigravity", ANTIGRAVITY_PUBLIC_MODELS, EXPECTED_LEADING_MODEL_ORDER],
    ["agy", AGY_PUBLIC_MODELS, EXPECTED_LEADING_MODEL_ORDER],
  ] as const) {
    assert.deepEqual(
@@ -114,7 +110,7 @@ test("Antigravity and AGY expose gemini-pro-agent and gemini-3.1-pro-high as cal
  );
 });

-test("Antigravity support catalogs expose every live Gemini 3.6 Flash tier", () => {
+test("Antigravity support catalogs expose every live Gemini 3.7 Flash tier", () => {
  const antigravityModelIds = new Set(ANTIGRAVITY_PUBLIC_MODELS.map((model) => model.id));
  const agyModelIds = new Set(AGY_PUBLIC_MODELS.map((model) => model.id));
  const cliAliases = new Set(CLI_TOOLS.antigravity.modelAliases);
@@ -123,7 +119,7 @@ test("Antigravity support catalogs expose every live Gemini 3.6 Flash tier", ()
    FREE_MODEL_BUDGETS.filter((model) => model.provider === "agy").map((model) => model.modelId)
  );

-  for (const modelId of CURRENT_36_FLASH_MODEL_IDS) {
+  for (const modelId of ACTIVE_FLASH_MODEL_IDS) {
    assert.equal(antigravityModelIds.has(modelId), true, `${modelId} missing from Antigravity`);
    assert.equal(agyModelIds.has(modelId), true, `${modelId} missing from AGY`);
    assert.equal(cliAliases.has(modelId), true, `${modelId} missing from CLI aliases`);
@@ -142,7 +138,7 @@ test("Antigravity support catalogs no longer advertise or price the rejected Hig
  assert.ok(pricing["gemini-pro-agent"]);
 });

-test("Antigravity and AGY support metadata excludes the retired Gemini 3 Flash id", () => {
+test("Antigravity and AGY support metadata excludes retired Flash ids", () => {
  const cliAliases = CLI_TOOLS.antigravity.modelAliases;
  const cliModelIds = CLI_TOOLS.antigravity.defaultModels.map((model) => model.id);
  const agyFreeModelIds = FREE_MODEL_BUDGETS.filter((model) => model.provider === "agy").map(
--- a/tests/unit/antigravity-thinking-config-preservation.test.ts
+++ b/tests/unit/antigravity-thinking-config-preservation.test.ts
@@ -10,7 +10,7 @@ import { FORMATS } from "../../open-sse/translator/formats.ts";

 test("antigravity reasoning capabilities: Gemini and Claude models support reasoning", () => {
  const geminiModels = [
-    "antigravity/gemini-3-flash-agent",
+    "antigravity/gemini-3.7-flash-high",
    "antigravity/gemini-pro-agent",
    "antigravity/gemini-3.1-pro-low",
    "antigravity/claude-sonnet-4-6",
--- a/tests/unit/antigravity-usage-service.test.ts
+++ b/tests/unit/antigravity-usage-service.test.ts
@@ -35,7 +35,7 @@ describe("getUsageForProvider (antigravity in usage.ts)", () => {
        ok: true,
        json: async () => ({
          models: {
-            "gemini-3-flash-agent": {
+            "gemini-3.7-flash-high": {
              quotaInfo: {
                remainingFraction: undefined,
                resetTime: "2026-05-26T00:00:00Z",
@@ -51,8 +51,8 @@ describe("getUsageForProvider (antigravity in usage.ts)", () => {
      assert.ok("quotas" in result, "should have quotas");

      if ("quotas" in result) {
-        const quota = result.quotas["gemini-3-flash-agent"];
-        assert.ok(quota, "should have quota for gemini-3-flash-agent");
+        const quota = result.quotas["gemini-3.7-flash-high"];
+        assert.ok(quota, "should have quota for gemini-3.7-flash-high");
        assert.equal(quota.remainingPercentage, 0, "remaining should be 0%");
        assert.equal(quota.unlimited, false, "should not be unlimited");
        assert.equal(quota.used > 0, true, "used should be > 0 when quota is exhausted");
@@ -69,7 +69,7 @@ describe("getUsageForProvider (antigravity in usage.ts)", () => {
        ok: true,
        json: async () => ({
          models: {
-            "gemini-3-flash-agent": {
+            "gemini-3.7-flash-high": {
              quotaInfo: {
                remainingFraction: 0,
                resetTime: "2026-05-26T00:00:00Z",
@@ -88,8 +88,8 @@ describe("getUsageForProvider (antigravity in usage.ts)", () => {
      assert.ok("quotas" in result, "should have quotas");

      if ("quotas" in result) {
-        const quota = result.quotas["gemini-3-flash-agent"];
-        assert.ok(quota, "should have quota for gemini-3-flash-agent");
+        const quota = result.quotas["gemini-3.7-flash-high"];
+        assert.ok(quota, "should have quota for gemini-3.7-flash-high");
        assert.equal(quota.remainingPercentage, 0, "remaining should be 0%");
        assert.equal(quota.unlimited, false, "should not be unlimited");
      }
@@ -105,7 +105,7 @@ describe("getUsageForProvider (antigravity in usage.ts)", () => {
        ok: true,
        json: async () => ({
          models: {
-            "gemini-3-flash-agent": {
+            "gemini-3.7-flash-high": {
              quotaInfo: {
                remainingFraction: 1.0,
                resetTime: "2026-05-26T00:00:00Z",
@@ -124,8 +124,8 @@ describe("getUsageForProvider (antigravity in usage.ts)", () => {
      assert.ok("quotas" in result, "should have quotas");

      if ("quotas" in result) {
-        const quota = result.quotas["gemini-3-flash-agent"];
-        assert.ok(quota, "should have quota for gemini-3-flash-agent");
+        const quota = result.quotas["gemini-3.7-flash-high"];
+        assert.ok(quota, "should have quota for gemini-3.7-flash-high");
        assert.equal(quota.remainingPercentage, 100, "remaining should be 100%");
        assert.equal(quota.unlimited, false, "should not be unlimited (has resetTime)");
      }
@@ -176,7 +176,7 @@ describe("getUsageForProvider (antigravity in usage.ts)", () => {
        ok: true,
        json: async () => ({
          models: {
-            "gemini-3-flash-agent": {
+            "gemini-3.7-flash-high": {
              quotaInfo: {
                remainingFraction: 0.5,
                resetTime: "2026-05-26T00:00:00Z",
@@ -195,8 +195,8 @@ describe("getUsageForProvider (antigravity in usage.ts)", () => {
      assert.ok("quotas" in result, "should have quotas");

      if ("quotas" in result) {
-        const quota = result.quotas["gemini-3-flash-agent"];
-        assert.ok(quota, "should have quota for gemini-3-flash-agent");
+        const quota = result.quotas["gemini-3.7-flash-high"];
+        assert.ok(quota, "should have quota for gemini-3.7-flash-high");
        assert.equal(quota.remainingPercentage, 50, "remaining should be 50%");
        assert.equal(quota.unlimited, false, "should not be unlimited");
      }
@@ -212,7 +212,7 @@ describe("getUsageForProvider (antigravity in usage.ts)", () => {
        ok: true,
        json: async () => ({
          models: {
-            "gemini-3-flash-agent": {
+            "gemini-3.7-flash-high": {
              quotaInfo: {
                remainingFraction: 1.5,
                resetTime: "2026-05-26T00:00:00Z",
@@ -231,8 +231,8 @@ describe("getUsageForProvider (antigravity in usage.ts)", () => {
      assert.ok("quotas" in result, "should have quotas");

      if ("quotas" in result) {
-        const quota = result.quotas["gemini-3-flash-agent"];
-        assert.ok(quota, "should have quota for gemini-3-flash-agent");
+        const quota = result.quotas["gemini-3.7-flash-high"];
+        assert.ok(quota, "should have quota for gemini-3.7-flash-high");
        assert.equal(quota.remainingPercentage, 100, "remaining should be clamped to 100%");
        assert.equal(quota.unlimited, false, "should not be unlimited (has resetTime)");
      }
@@ -248,7 +248,7 @@ describe("getUsageForProvider (antigravity in usage.ts)", () => {
        ok: true,
        json: async () => ({
          models: {
-            "gemini-3-flash-agent": {
+            "gemini-3.7-flash-high": {
              quotaInfo: {
                remainingFraction: -0.5,
                resetTime: "2026-05-26T00:00:00Z",
@@ -267,8 +267,8 @@ describe("getUsageForProvider (antigravity in usage.ts)", () => {
      assert.ok("quotas" in result, "should have quotas");

      if ("quotas" in result) {
-        const quota = result.quotas["gemini-3-flash-agent"];
-        assert.ok(quota, "should have quota for gemini-3-flash-agent");
+        const quota = result.quotas["gemini-3.7-flash-high"];
+        assert.ok(quota, "should have quota for gemini-3.7-flash-high");
        assert.equal(quota.remainingPercentage, 0, "remaining should be clamped to 0%");
        assert.equal(quota.unlimited, false, "should not be unlimited");
      }
--- a/tests/unit/antigravity-weekly-quota-4017.test.ts
+++ b/tests/unit/antigravity-weekly-quota-4017.test.ts
@@ -176,7 +176,7 @@ test("getUsageForProvider(antigravity) merges weekly quotas with the selected CL
        json: async () => ({
          buckets: [
            {
-              modelId: "gemini-3-flash-agent",
+              modelId: "gemini-3.7-flash-high",
              remainingFraction: 0.4,
              resetTime: RESET_IN_2_HOURS,
            },
@@ -190,7 +190,7 @@ test("getUsageForProvider(antigravity) merges weekly quotas with the selected CL
      ok: true,
      json: async () => ({
        models: {
-          "gemini-3-flash-agent": {
+          "gemini-3.7-flash-high": {
            quotaInfo: { remainingFraction: 1.0, resetTime: RESET_IN_2_HOURS },
          },
        },
@@ -211,8 +211,8 @@ test("getUsageForProvider(antigravity) merges weekly quotas with the selected CL
  const quotas = (result as UsageResult).quotas;

  // Existing per-model 5h quota is untouched.
-  assert.ok(quotas["gemini-3-flash-agent"], "per-model 5h quota still present");
-  assert.equal(quotas["gemini-3-flash-agent"].quotaSource, "retrieveUserQuota");
+  assert.ok(quotas["gemini-3.7-flash-high"], "per-model 5h quota still present");
+  assert.equal(quotas["gemini-3.7-flash-high"].quotaSource, "retrieveUserQuota");

  // New weekly group quota is merged in alongside it.
  assert.ok(quotas.gemini_weekly, "weekly group quota merged in");
@@ -235,7 +235,7 @@ test("getUsageForProvider(antigravity) is unaffected when retrieveUserQuotaSumma
      ok: true,
      json: async () => ({
        models: {
-          "gemini-3-flash-agent": {
+          "gemini-3.7-flash-high": {
            quotaInfo: { remainingFraction: 1.0, resetTime: RESET_IN_2_HOURS },
          },
        },
@@ -253,6 +253,6 @@ test("getUsageForProvider(antigravity) is unaffected when retrieveUserQuotaSumma

  const result = await getUsageForProvider(connection, { forceRefresh: true });
  const quotas = (result as UsageResult).quotas;
-  assert.ok(quotas["gemini-3-flash-agent"], "per-model quota still present without weekly data");
+  assert.ok(quotas["gemini-3.7-flash-high"], "per-model quota still present without weekly data");
  assert.equal(quotas.gemini_weekly, undefined, "no weekly key when the RPC is unavailable");
 });
--- a/tests/unit/auto-best-free-tier-filter.test.ts
+++ b/tests/unit/auto-best-free-tier-filter.test.ts
@@ -11,7 +11,7 @@
 * `chat.ts` routes through `resolveAutoRoutingState` → `createVirtualAutoCombo`
 * (autoRouting.ts), NOT through `createBuiltinAutoCombo`. So the chat path
 * skipped the tier filter entirely and `auto/best-free` behaved as plain
- * `auto/cheap`, allowing paid models (e.g. antigravity/gemini-3.6-flash-high)
+ * `auto/cheap`, allowing paid models (e.g. antigravity/gemini-3.7-flash-high)
 * to be selected from the full pool.
 *
 * classifyAutoModel() is module-private, so this exercises it through the public
--- a/tests/unit/auto-combo-credentialed-model-pool.test.ts
+++ b/tests/unit/auto-combo-credentialed-model-pool.test.ts
@@ -87,9 +87,9 @@ test("credentialed providers expose one logical candidate per visible registry m
  );
  for (const model of [
    "antigravity/claude-sonnet-4-6",
-    "antigravity/gemini-3.6-flash-low",
-    "antigravity/gemini-3.6-flash-medium",
-    "antigravity/gemini-3.6-flash-high",
+    "antigravity/gemini-3.7-flash-low",
+    "antigravity/gemini-3.7-flash-medium",
+    "antigravity/gemini-3.7-flash-high",
  ]) {
    assert.ok(modelStrings.includes(model), `${model} should be eligible for auto routing`);
  }
@@ -121,15 +121,15 @@ test("candidate transparency expands a logical model into per-account rows", asy
 });

 test("connection model exclusions narrow only that model's account allowlist", async () => {
-  const { first, second } = await seedConnections(["gemini-3.6-*"]);
+  const { first, second } = await seedConnections(["gemini-3.7-*"]);

  const combo = await virtualFactory.createVirtualAutoCombo(undefined);
  const candidates = antigravityCandidates(combo);
  const geminiCandidates = candidates.filter((candidate) =>
-    candidate.model.startsWith("antigravity/gemini-3.6-")
+    candidate.model.startsWith("antigravity/gemini-3.7-")
  );

-  assert.ok(geminiCandidates.length >= 3, "Gemini 3.6 candidates should remain available");
+  assert.ok(geminiCandidates.length >= 3, "Gemini 3.7 candidates should remain available");
  for (const candidate of geminiCandidates) {
    assert.deepEqual(candidate.allowedConnectionIds, [second.id]);
  }
@@ -150,9 +150,9 @@ test("hiding the first registry model does not drop the credentialed provider",

  assert.equal(modelStrings.includes("antigravity/claude-sonnet-4-6"), false);
  for (const model of [
-    "antigravity/gemini-3.6-flash-low",
-    "antigravity/gemini-3.6-flash-medium",
-    "antigravity/gemini-3.6-flash-high",
+    "antigravity/gemini-3.7-flash-low",
+    "antigravity/gemini-3.7-flash-medium",
+    "antigravity/gemini-3.7-flash-high",
  ]) {
    assert.ok(modelStrings.includes(model), `${model} should remain after Sonnet is hidden`);
  }
--- a/tests/unit/chat-messages-validation-6402.test.ts
+++ b/tests/unit/chat-messages-validation-6402.test.ts
@@ -181,16 +181,10 @@ test("Antigravity Gemini-family regression covers every current callable tier",
  for (const model of [
    "gemini-pro-agent",
    "gemini-3.1-pro-low",
-    "gemini-3.6-flash-high",
-    "gemini-3.6-flash-medium",
-    "gemini-3.6-flash-low",
-    "gemini-3-flash-agent",
-    "gemini-3.5-flash-low",
-    "gemini-3.5-flash-extra-low",
+    "gemini-3.7-flash-high",
+    "gemini-3.7-flash-medium",
+    "gemini-3.7-flash-low",
    "gemini-3.1-flash-lite",
-    "gemini-2.5-flash-thinking",
-    "gemini-2.5-flash",
-    "gemini-2.5-flash-lite",
  ]) {
    assert.ok(coveredModels.has(model), `expected regression coverage for ${model}`);
  }
--- a/tests/unit/combo-prescreen.test.ts
+++ b/tests/unit/combo-prescreen.test.ts
@@ -210,7 +210,7 @@ test("priority combo: quota 429 on passthrough provider does not skip another mo
  const combo = await combosDb.createCombo({
    name: "passthrough-quota-scope",
    strategy: "priority",
-    models: ["antigravity/claude-opus-4-6-thinking", "antigravity/gemini-3-flash-agent"],
+    models: ["antigravity/claude-opus-4-6-thinking", "antigravity/gemini-3.7-flash-high"],
  });

  const response = await handleComboChat({
@@ -235,7 +235,7 @@ test("priority combo: quota 429 on passthrough provider does not skip another mo
  });

  assert.equal(response.status, 200);
-  assert.equal(calls.at(-1), "antigravity/gemini-3-flash-agent");
+  assert.equal(calls.at(-1), "antigravity/gemini-3.7-flash-high");
  assert.ok(
    calls.includes("antigravity/claude-opus-4-6-thinking"),
    "first passthrough model should be attempted before fallback"
--- a/tests/unit/devin-cli-catalog.test.ts
+++ b/tests/unit/devin-cli-catalog.test.ts
@@ -28,10 +28,10 @@ test("Devin CLI catalog includes the refreshed native model ids", () => {
    "claude-opus-5-high",
    "claude-opus-5-xhigh",
    "claude-opus-5-max",
-    "gemini-3-6-flash-minimal",
-    "gemini-3-6-flash-low",
-    "gemini-3-6-flash-medium",
-    "gemini-3-6-flash-high",
+    "gemini-3-7-flash-minimal",
+    "gemini-3-7-flash-low",
+    "gemini-3-7-flash-medium",
+    "gemini-3-7-flash-high",
    "kimi-k3-low",
    "kimi-k3-high",
    "kimi-k3-max",
--- a/tests/unit/executor-agy.test.ts
+++ b/tests/unit/executor-agy.test.ts
@@ -29,7 +29,7 @@ test("getExecutor('antigravity') returns AntigravityExecutor", () => {

 test("getExecutor('agy') builds valid streaming URL", () => {
  const executor = getExecutor("agy");
-  const url = executor.buildUrl("gemini-3.5-flash-high", true);
+  const url = executor.buildUrl("gemini-3.7-flash-high", true);
  assert.ok(
    url.includes("streamGenerateContent?alt=sse"),
    `expected streaming endpoint URL, got: ${url}`
@@ -38,7 +38,7 @@ test("getExecutor('agy') builds valid streaming URL", () => {

 test("getExecutor('agy') builds valid non-streaming URL", () => {
  const executor = getExecutor("agy");
-  const url = executor.buildUrl("gemini-3.5-flash-high", false);
+  const url = executor.buildUrl("gemini-3.7-flash-high", false);
  // Antigravity executor always uses streaming endpoint (buildUrl ignores stream flag)
  assert.ok(
    url.includes("streamGenerateContent?alt=sse"),
--- a/tests/unit/executor-antigravity.test.ts
+++ b/tests/unit/executor-antigravity.test.ts
@@ -516,7 +516,7 @@ test("AntigravityExecutor.collectStreamToResponse converts textual tool call SSE

  const result = await executor.collectStreamToResponse(
    response,
-    "gemini-3.5-flash-low",
+    "gemini-3.7-flash-low",
    "https://example.com",
    { Authorization: "Bearer ag-token" },
    { request: {} }
--- a/tests/unit/executor-notion-web.test.ts
+++ b/tests/unit/executor-notion-web.test.ts
@@ -48,11 +48,27 @@ describe("NotionWebExecutor — registry consistency", () => {

  it("registers a model catalog reachable via getModelsByProviderId", () => {
    const models = getModelsByProviderId("notion-web");
-    assert.ok(models.length >= 1);
-    assert.ok(models.some((m) => m.id === "notion-ai"));
-    // Seed catalog uses real web-picker labels (fable-5 / gpt-5.6-sol), not food codenames.
-    assert.ok(
-      models.some((m) => m.id === "fable-5" || m.id === "gpt-5.6-sol" || m.id === "opus-4.8")
+    assert.deepEqual(
+      models.map((m) => m.id),
+      [
+        "notion-ai",
+        "gpt-5.6-sol",
+        "gpt-5.6-terra",
+        "gpt-5.6-luna",
+        "gpt-5.4-mini",
+        "gpt-5.4-nano",
+        "gemini-3.7-flash",
+        "gemini-3.1-pro",
+        "fable-5",
+        "opus-5",
+        "sonnet-5",
+        "haiku-4.5",
+        "grok-4.6",
+        "kimi-k3",
+        "kimi-k2.7-code",
+        "deepseek-v4-pro",
+        "glm-5.2",
+      ]
    );
    assert.equal(
      models.some(
@@ -103,6 +119,7 @@ const COOKIE_WITH_SPACE = "token_v2=xyz; space_id=space-1; notion_user_id=user-1

 describe("NotionWebExecutor — upstream translation (mocked TLS fetch)", () => {
  it("posts createThread + config/context/user and returns a chat.completion", async () => {
+    mod.__resetNotionThreadSessionsForTests();
    const executor = new mod.NotionWebExecutor();
    let capturedUrl = "";
    let capturedHeaders: Record<string, string> = {};
--- a/tests/unit/executor-promptql.test.ts
+++ b/tests/unit/executor-promptql.test.ts
@@ -213,7 +213,7 @@ describe("PromptQl — helpers", () => {
    assert.equal(models.clientFacingPromptQlModelId("pql/gpt-5.6-sol"), "gpt-5.6-sol");
    const r = models.resolvePromptQlModel("Claude Fable 5");
    assert.ok(r);
-    assert.equal(r!.id, "vertex-claude-fable-5");
+    assert.equal(r!.id, "bedrock-claude-fable-5");
  });

  it("converts credit micros to USD", () => {
--- a/tests/unit/gemini-imagen-predict.test.ts
+++ b/tests/unit/gemini-imagen-predict.test.ts
@@ -1,86 +0,0 @@
-/**
- * Google AI Studio (Gemini API) Imagen support on /v1/images/generations.
- *
- * Imagen uses the dedicated ":predict" endpoint (instances/parameters body,
- * base64 predictions), NOT generateContent. Before this, `gemini/imagen-4.0-*`
- * was advertised in /v1/models but unroutable — the image route rejected it with
- * "Invalid image model" because `gemini` was not in the image registry.
- *
- * These cover the pure request-builder / response-parser and the registry wiring.
- * The live Google call is not exercised (Imagen needs a billing-enabled key).
- */
-import test from "node:test";
-import assert from "node:assert/strict";
-
-import { IMAGE_PROVIDERS, parseImageModel } from "../../open-sse/config/imageRegistry.ts";
-import {
-  buildImagenPredictBody,
-  parseImagenPredictResponse,
-  isImagenModel,
-} from "../../open-sse/handlers/imageGeneration/providers/googleImagen.ts";
-
-test("gemini image provider is registered for the Imagen family via google-imagen format", () => {
-  const gemini = IMAGE_PROVIDERS.gemini;
-  assert.ok(gemini, "gemini image provider must exist");
-  assert.equal(gemini.format, "google-imagen");
-  assert.equal(gemini.authHeader, "x-goog-api-key");
-  assert.equal(gemini.baseUrl, "https://generativelanguage.googleapis.com/v1beta/models");
-  assert.deepEqual(
-    gemini.models.map((m) => m.id),
-    ["imagen-4.0-generate-001", "imagen-4.0-ultra-generate-001", "imagen-4.0-fast-generate-001"]
-  );
-});
-
-test("parseImageModel resolves gemini/imagen-4.0-* to the gemini provider", () => {
-  assert.deepEqual(parseImageModel("gemini/imagen-4.0-generate-001"), {
-    provider: "gemini",
-    model: "imagen-4.0-generate-001",
-  });
-});
-
-test("isImagenModel gates only the Imagen family (flash-image belongs on the chat route)", () => {
-  assert.equal(isImagenModel("imagen-4.0-generate-001"), true);
-  assert.equal(isImagenModel("imagen-4.0-ultra-generate-001"), true);
-  assert.equal(isImagenModel("gemini-2.5-flash-image"), false);
-  assert.equal(isImagenModel("nano-banana-pro"), false);
-  assert.equal(isImagenModel(""), false);
-  assert.equal(isImagenModel(undefined), false);
-});
-
-test("buildImagenPredictBody produces the :predict instances/parameters shape", () => {
-  const body = buildImagenPredictBody({ prompt: "a red apple", n: 2, size: "1792x1024" });
-  assert.deepEqual(body, {
-    instances: [{ prompt: "a red apple" }],
-    parameters: { sampleCount: 2, aspectRatio: "16:9" },
-  });
-});
-
-test("buildImagenPredictBody clamps sampleCount to [1,4] and defaults aspectRatio to 1:1", () => {
-  assert.equal(buildImagenPredictBody({ prompt: "x" }).parameters.sampleCount, 1);
-  assert.equal(buildImagenPredictBody({ prompt: "x", n: 0 }).parameters.sampleCount, 1);
-  assert.equal(buildImagenPredictBody({ prompt: "x", n: 99 }).parameters.sampleCount, 4);
-  assert.equal(buildImagenPredictBody({ prompt: "x" }).parameters.aspectRatio, "1:1");
-  // Native aspect ratio passes through.
-  assert.equal(buildImagenPredictBody({ prompt: "x", aspect_ratio: "9:16" }).parameters.aspectRatio, "9:16");
-});
-
-test("parseImagenPredictResponse normalizes predictions[].bytesBase64Encoded to OpenAI shape", () => {
-  const out = parseImagenPredictResponse(
-    {
-      predictions: [
-        { bytesBase64Encoded: "AAAA", mimeType: "image/png" },
-        { bytesBase64Encoded: "BBBB", mimeType: "image/png" },
-      ],
-    },
-    "a red apple"
-  );
-  assert.equal(out.data.length, 2);
-  assert.deepEqual(out.data[0], { b64_json: "AAAA", revised_prompt: "a red apple" });
-  assert.equal(typeof out.created, "number");
-});
-
-test("parseImagenPredictResponse tolerates empty/absent predictions", () => {
-  assert.deepEqual(parseImagenPredictResponse({}, "x").data, []);
-  assert.deepEqual(parseImagenPredictResponse({ predictions: [] }, "x").data, []);
-  assert.deepEqual(parseImagenPredictResponse({ predictions: [{}] }, "x").data, []);
-});
--- a/tests/unit/gemini-models-parser.test.ts
+++ b/tests/unit/gemini-models-parser.test.ts
@@ -3,9 +3,7 @@ import assert from "node:assert/strict";

 import { parseGeminiModelsList } from "../../src/lib/providerModels/geminiModelsParser";

-// A representative slice of the live generativelanguage v1beta/models response — including the
-// image models (gemini-*-image via generateContent, imagen-* via predict) that the Vertex catalog
-// must surface dynamically.
+// A representative slice of the live generativelanguage v1beta/models response.
 const SAMPLE = {
  models: [
    {
@@ -21,11 +19,6 @@ const SAMPLE = {
      displayName: "Gemini 3 Pro Image Preview",
      supportedGenerationMethods: ["generateContent", "countTokens"],
    },
-    {
-      name: "models/imagen-4.0-generate-001",
-      displayName: "Imagen 4.0",
-      supportedGenerationMethods: ["predict"],
-    },
    {
      name: "models/text-embedding-004",
      displayName: "Text Embedding 004",
@@ -41,13 +34,6 @@ const SAMPLE = {
      displayName: "Veo 3.0",
      supportedGenerationMethods: ["predictLongRunning"],
    },
-    {
-      // Defensive: an Imagen model exposed via a long-running method must stay
-      // "images", never "video".
-      name: "models/imagen-future-preview",
-      displayName: "Imagen Future",
-      supportedGenerationMethods: ["predictLongRunning"],
-    },
  ],
 };

@@ -70,13 +56,6 @@ test("parseGeminiModelsList maps generateContent image models to the chat endpoi
  assert.deepEqual(proImage!.supportedEndpoints, ["chat"]);
 });

-test("parseGeminiModelsList maps Imagen predict models to the images endpoint", () => {
-  const models = parseGeminiModelsList(SAMPLE);
-  const imagen = models.find((m) => m.id === "imagen-4.0-generate-001");
-  assert.ok(imagen, "imagen-4.0-generate-001 should be present");
-  assert.deepEqual(imagen!.supportedEndpoints, ["images"]);
-});
-
 test("parseGeminiModelsList maps embedContent and bidiGenerateContent", () => {
  const models = parseGeminiModelsList(SAMPLE);
  assert.deepEqual(models.find((m) => m.id === "text-embedding-004")!.supportedEndpoints, [
@@ -94,13 +73,6 @@ test("parseGeminiModelsList maps Veo predictLongRunning models to the video endp
  assert.deepEqual(veo!.supportedEndpoints, ["video"]);
 });

-test("parseGeminiModelsList keeps Imagen as images even via a long-running method", () => {
-  const models = parseGeminiModelsList(SAMPLE);
-  const imagen = models.find((m) => m.id === "imagen-future-preview");
-  assert.ok(imagen, "imagen-future-preview should be present");
-  assert.deepEqual(imagen!.supportedEndpoints, ["images"]);
-});
-
 test("parseGeminiModelsList defaults to chat and tolerates empty/missing input", () => {
  assert.deepEqual(parseGeminiModelsList({}), []);
  assert.deepEqual(parseGeminiModelsList(null), []);
--- a/tests/unit/gemini-web.test.ts
+++ b/tests/unit/gemini-web.test.ts
@@ -171,7 +171,7 @@ test("Provider: gemini-web has correct models", async () => {
    models.map((m: any) => [m.id, m.name]),
    [
      ["gemini-3.1-pro", "Gemini 3.1 Pro"],
-      ["gemini-3.5-flash", "Gemini 3.5 Flash"],
+      ["gemini-3.7-flash", "Gemini 3.7 Flash"],
      ["gemini-3.1-flash-lite", "Gemini 3.1 Flash-Lite"],
    ]
  );
--- a/tests/unit/ghe-copilot-targetformat-parity.test.ts
+++ b/tests/unit/ghe-copilot-targetformat-parity.test.ts
@@ -8,7 +8,8 @@
 import test from "node:test";
 import assert from "node:assert/strict";

-const { gheCopilotProvider } = await import("../../open-sse/config/providers/registry/ghe-copilot/index.ts");
+const { gheCopilotProvider } =
+  await import("../../open-sse/config/providers/registry/ghe-copilot/index.ts");
 const { getModelsByProviderId } = await import("../../open-sse/config/providerModels.ts");

 type ModelEntry = { id: string; targetFormat?: string; [k: string]: unknown };
@@ -30,7 +31,7 @@ const MUST_NOT_BE_RESPONSES = [
  "claude-sonnet-4.5",
  "claude-haiku-4.5",
  "gemini-3.1-pro-preview",
-  "gemini-3.5-flash",
+  "gemini-3.7-flash",
 ];

 for (const id of MUST_NOT_BE_RESPONSES) {
--- a/tests/unit/github-copilot-model-discovery.test.ts
+++ b/tests/unit/github-copilot-model-discovery.test.ts
@@ -140,7 +140,7 @@ test("curated Copilot allowlist contains the final approved model ids only", ()
      "claude-sonnet-4.5",
      "claude-haiku-4.5",
      "gemini-3.1-pro-preview",
-      "gemini-3.5-flash",
+      "gemini-3.7-flash",
      "gpt-5.6-sol",
      "gpt-5.6-terra",
      "gpt-5.6-luna",
--- a/tests/unit/model-capabilities-registry.test.ts
+++ b/tests/unit/model-capabilities-registry.test.ts
@@ -154,7 +154,7 @@ test("unknown models keep maxOutputTokens null instead of using a generic defaul
  );
 });

-test("Antigravity Gemini 3.5 upstream IDs share the Flash capability profile", () => {
+test("provider-neutral Gemini 3.5 tier IDs retain their non-thinking capabilities", () => {
  for (const modelId of [
    "gemini-3.5-flash-extra-low",
    "gemini-3.5-flash-low",
@@ -162,22 +162,21 @@ test("Antigravity Gemini 3.5 upstream IDs share the Flash capability profile", (
  ]) {
    const spec = MODEL_SPECS[modelId];
    assert.ok(spec, `missing exact MODEL_SPECS entry for ${modelId}`);
-    const capabilities = modelCapabilities.getResolvedModelCapabilities(`antigravity/${modelId}`);
+    const capabilities = modelCapabilities.getResolvedModelCapabilities(modelId);
    assert.equal(capabilities.contextWindow, 1048576, modelId);
    assert.equal(capabilities.maxOutputTokens, 65536, modelId);
-    assert.equal(capabilities.supportsThinking, true, modelId);
+    // These ids encode the upstream reasoning tier and do not accept a client-supplied effort.
+    assert.equal(capabilities.supportsThinking, false, modelId);
    assert.equal(capabilities.supportsTools, true, modelId);
    assert.equal(capabilities.supportsVision, true, modelId);
  }
 });

-test("Antigravity Gemini 3.7 and 3.6 tier IDs share the Flash capability profile", () => {
+test("Antigravity Gemini 3.7 tier IDs share the Flash capability profile", () => {
  for (const modelId of [
    "gemini-3.7-flash-high",
    "gemini-3.7-flash-medium",
-    "gemini-3.6-flash-high",
-    "gemini-3.6-flash-medium",
-    "gemini-3.6-flash-low",
+    "gemini-3.7-flash-low",
  ]) {
    const spec = MODEL_SPECS[modelId];
    assert.ok(spec, `missing exact MODEL_SPECS entry for ${modelId}`);
--- a/tests/unit/models-catalog-route.test.ts
+++ b/tests/unit/models-catalog-route.test.ts
@@ -696,12 +696,15 @@ test("v1 models catalog exposes current Antigravity aliases without retired mode
  assert.equal(ids.has("antigravity/gemini-3.1-pro"), false);
  assert.equal(ids.has("antigravity/gemini-2.5-computer-use-preview-10-2025"), false);
  assert.equal(ids.has("antigravity/rev19-uic3-1p"), false);
-  assert.ok(ids.has("antigravity/gemini-3.6-flash-high"));
-  assert.ok(ids.has("antigravity/gemini-3.6-flash-medium"));
-  assert.ok(ids.has("antigravity/gemini-3.6-flash-low"));
-  assert.ok(ids.has("antigravity/gemini-3.5-flash-extra-low"));
-  assert.ok(ids.has("antigravity/gemini-3.5-flash-low"));
-  assert.ok(ids.has("antigravity/gemini-3-flash-agent"));
+  assert.ok(ids.has("antigravity/gemini-3.7-flash-high"));
+  assert.ok(ids.has("antigravity/gemini-3.7-flash-medium"));
+  assert.ok(ids.has("antigravity/gemini-3.7-flash-low"));
+  assert.equal(ids.has("antigravity/gemini-3.6-flash-high"), false);
+  assert.equal(ids.has("antigravity/gemini-3.6-flash-medium"), false);
+  assert.equal(ids.has("antigravity/gemini-3.6-flash-low"), false);
+  assert.equal(ids.has("antigravity/gemini-3.5-flash-extra-low"), false);
+  assert.equal(ids.has("antigravity/gemini-3.5-flash-low"), false);
+  assert.equal(ids.has("antigravity/gemini-3-flash-agent"), false);
  assert.equal(ids.has("antigravity/gemini-3.5-flash-medium"), false);
  assert.equal(ids.has("antigravity/gemini-3.5-flash-high"), false);
  assert.equal(ids.has("antigravity/gemini-3.5-flash-preview"), false);
--- a/tests/unit/opencode-go-effort-aliases-8353.test.ts
+++ b/tests/unit/opencode-go-effort-aliases-8353.test.ts
@@ -98,7 +98,12 @@ test("#8353 catalog: aliases are NOT synthesized on opencode-zen", () => {
  for (const { alias } of ISSUE_ALIASES) {
    assert.equal(zenIds.has(alias), false, `opencode-zen must not expose ${alias}`);
  }
+  // kimi-k3 became a live zen model in the 2026-08-17 registry sync (present in
+  // https://opencode.ai/zen/v1/models) — only the remaining Go-tier-only bases
+  // must stay absent from zen. The effort alias kimi-k3-max is still Go-only
+  // and remains covered by the ISSUE_ALIASES loop above.
  for (const base of NEW_BASES) {
+    if (base === "kimi-k3") continue;
    assert.equal(zenIds.has(base), false, `opencode-zen must not expose base ${base}`);
  }
 });
--- a/tests/unit/perplexity-web-model-mappings.test.ts
+++ b/tests/unit/perplexity-web-model-mappings.test.ts
@@ -27,9 +27,14 @@ test("Perplexity Web registers the refreshed model catalog", () => {
    "pplx-sonar",
    "pplx-kimi",
    "pplx-glm",
-    "pplx-grok-4.5",
+    "pplx-grok-4.6",
  ];
  assert.deepEqual([...modelIds].sort(), expectedModelIds.sort());
+
+  const modelNames = new Map(models.map((model) => [model.id, model.name]));
+  assert.equal(modelNames.get("pplx-gemini"), "Gemini 3.7 Flash (via Perplexity)");
+  assert.equal(modelNames.get("pplx-kimi"), "Kimi K3 (via Perplexity)");
+  assert.equal(modelNames.get("pplx-grok-4.6"), "Grok 4.6 (via Perplexity)");
 });

 test("every advertised Perplexity Web model has an explicit internal mapping", () => {
@@ -37,10 +42,16 @@ test("every advertised Perplexity Web model has an explicit internal mapping", (
  assert.deepEqual(missing, []);
  assert.deepEqual(MODEL_MAP["pplx-gpt-5.6-terra"], ["copilot", "gpt56_terra"]);
  assert.deepEqual(MODEL_MAP["pplx-gpt-5.6-sol"], ["copilot", "gpt56_sol"]);
-  assert.deepEqual(MODEL_MAP["pplx-grok-4.5"], ["copilot", "grok45low"]);
+  assert.deepEqual(MODEL_MAP["pplx-gemini"], ["copilot", "gemini37flash"]);
+  assert.deepEqual(MODEL_MAP["pplx-kimi"], ["copilot", "kimik3thinking"]);
+  assert.deepEqual(MODEL_MAP["pplx-grok-4.6"], ["copilot", "grok46low"]);
  assert.deepEqual(MODEL_MAP["pplx-opus"], ["copilot", "claude50opus"]);
+  assert.equal(THINKING_MAP["pplx-gemini"], "gemini37flashthinking");
  assert.equal(THINKING_MAP["pplx-opus"], "claude50opusthinking");
  assert.equal(THINKING_MAP["pplx-gpt-5.6-terra"], "gpt56_terra_thinking");
  assert.equal(THINKING_MAP["pplx-gpt-5.6-sol"], "gpt56_sol_thinking");
-  assert.equal(THINKING_MAP["pplx-grok-4.5"], "grok45medium");
+  assert.equal(THINKING_MAP["pplx-kimi"], "kimik3thinking");
+  assert.equal(THINKING_MAP["pplx-grok-4.6"], "grok46medium");
+  assert.equal(MODEL_MAP["pplx-grok-4.5"], undefined);
+  assert.equal(THINKING_MAP["pplx-grok-4.5"], undefined);
 });
--- a/tests/unit/pricing-ag-flash-tiers.test.ts
+++ b/tests/unit/pricing-ag-flash-tiers.test.ts
@@ -3,55 +3,52 @@ import assert from "node:assert/strict";

 import { getDefaultPricing } from "../../src/shared/constants/pricing.ts";

-// Antigravity exposes Gemini 3.5 Flash via three public client IDs in
-// ANTIGRAVITY_PUBLIC_MODELS (`open-sse/config/antigravityModelAliases.ts`):
-//   - gemini-3.5-flash-extra-low → "Gemini 3.5 Flash (Low)" — upstream Low tier
-//   - gemini-3-flash-agent   → "Gemini 3.5 Flash (High)"   — upstream High tier
-//   - gemini-3.5-flash-low   → "Gemini 3.5 Flash (Medium)" — upstream Medium tier
-//   - gemini-pro-agent       → "Gemini 3.1 Pro (High)"     — upstream Pro High alias
-// All three were missing pricing rows in `ag` (DEFAULT_PRICING.ag), so
-// getPricingForModel("ag", id) returned null and downstream cost / quota
-// calculations silently fell back to $0. Each row matches its upstream quota tier.
+const EXPECTED_GEMINI_3_7_PROMO_PRICING = {
+  input: 0.75,
+  output: 3.75,
+  cached: 0.075,
+  reasoning: 3.75,
+  cache_creation: 0.75,
+};

-for (const [modelId, tier] of [
-  ["gemini-3.5-flash-extra-low", "Low"],
-  ["gemini-3.5-flash-low", "Medium"],
-  ["gemini-3-flash-agent", "High"],
-] as const) {
-  test(`ag/${modelId} matches the Gemini 3.5 Flash (${tier}) tier`, () => {
-    const p = getDefaultPricing().ag[modelId];
-    assert.ok(p);
-    assert.equal(p.input, 0.5);
-    assert.equal(p.output, 3.0);
-    assert.equal(p.cached, 0.03);
-    assert.equal(p.reasoning, 4.5);
-    assert.equal(p.cache_creation, 0.5);
-  });
-}
+const ACTIVE_FLASH_IDS = [
+  "gemini-3.7-flash-low",
+  "gemini-3.7-flash-medium",
+  "gemini-3.7-flash-high",
+] as const;

-test("ag/gemini-pro-agent matches the Gemini 3.1 Pro (High) tier", () => {
-  const p = getDefaultPricing().ag["gemini-pro-agent"];
-  assert.equal(p.input, 4.0);
-  assert.equal(p.output, 18.0);
-  assert.equal(p.cached, 0.5);
-  assert.equal(p.reasoning, 27.0);
-  assert.equal(p.cache_creation, 4.0);
+const RETIRED_ANTIGRAVITY_FLASH_IDS = [
+  "gemini-3.6-flash-low",
+  "gemini-3.6-flash-medium",
+  "gemini-3.6-flash-high",
+  "gemini-3.5-flash-extra-low",
+  "gemini-3.5-flash-low",
+  "gemini-3.5-flash-medium",
+  "gemini-3.5-flash-high",
+  "gemini-3-flash-agent",
+] as const;
+
+test("Antigravity provider namespaces price every live Gemini 3.7 Flash tier", () => {
+  const pricing = getDefaultPricing() as Record<string, Record<string, unknown>>;
+
+  for (const provider of ["ag", "antigravity", "agy"] as const) {
+    for (const modelId of ACTIVE_FLASH_IDS) {
+      assert.deepEqual(pricing[provider][modelId], EXPECTED_GEMINI_3_7_PROMO_PRICING);
+    }
+  }
 });

-// Gemini 3.6 Flash (released 2026-07-21) ships three public client IDs in
-// ANTIGRAVITY_PUBLIC_MODELS (`open-sse/config/antigravityModelAliases.ts`) and
-// MODEL_SPECS (`src/shared/constants/modelSpecs.ts`), but was missing pricing
-// rows in `ag` (DEFAULT_PRICING.ag) — cost / quota calculations silently fell
-// back to $0. Pricing: $1.50 input / $7.50 output / $0.15 cached per MTok
-// (Google's 2026-07-21 announcement). Thinking tokens billed at output rate.
-for (const tier of ["low", "medium", "high"]) {
-  test(`ag/gemini-3.6-flash-${tier} has a non-null pricing row`, () => {
-    const p = getDefaultPricing().ag[`gemini-3.6-flash-${tier}`];
-    assert.ok(p, `expected a pricing row for ag/gemini-3.6-flash-${tier}`);
-    assert.equal(p.input, 1.5);
-    assert.equal(p.output, 7.5);
-    assert.equal(p.cached, 0.15);
-    assert.equal(p.reasoning, 7.5);
-    assert.equal(p.cache_creation, 1.5);
-  });
-}
+test("Antigravity pricing excludes retired Flash ids without affecting other providers", () => {
+  const pricing = getDefaultPricing() as Record<string, Record<string, unknown>>;
+
+  for (const provider of ["ag", "antigravity", "agy"] as const) {
+    for (const modelId of RETIRED_ANTIGRAVITY_FLASH_IDS) {
+      assert.equal(pricing[provider][modelId], undefined, `${provider}/${modelId}`);
+    }
+  }
+
+  assert.ok(
+    pricing.orcarouter["google/gemini-3.6-flash"],
+    "OrcaRouter must retain the Gemini 3.6 model it still serves"
+  );
+});
--- a/tests/unit/provider-columns.test.ts
+++ b/tests/unit/provider-columns.test.ts
@@ -87,9 +87,9 @@ test("getProviderColumns: Antigravity falls back to dynamic schema (first 3 quot
      "claude-opus-4-6-thinking": { used: 0, total: 100, remainingPercentage: 100 },
      "claude-sonnet-4-6": { used: 0, total: 100, remainingPercentage: 100 },
      "gemini-3.1-pro-low": { used: 0, total: 100, remainingPercentage: 100 },
-      "gemini-3.5-flash-low": { used: 0, total: 100, remainingPercentage: 100 },
-      "gemini-3.5-flash-medium": { used: 0, total: 100, remainingPercentage: 100 },
-      "gemini-3.5-flash-high": { used: 0, total: 100, remainingPercentage: 100 },
+      "gemini-3.7-flash-low": { used: 0, total: 100, remainingPercentage: 100 },
+      "gemini-3.7-flash-medium": { used: 0, total: 100, remainingPercentage: 100 },
+      "gemini-3.7-flash-high": { used: 0, total: 100, remainingPercentage: 100 },
    },
  });

--- a/tests/unit/provider-models-config.test.ts
+++ b/tests/unit/provider-models-config.test.ts
@@ -99,7 +99,7 @@ test("GitHub Copilot registry reflects the current supported model lineup", () =
  // the OpenAI shape. Port of decolua/9router#2608.
  assert.equal(getModelTargetFormat("gh", "claude-opus-4.8-fast"), "claude");
  assert.equal(getModelTargetFormat("gh", "claude-sonnet-4.6"), "claude");
-  assert.equal(getModelTargetFormat("gh", "gemini-3.5-flash"), null);
+  assert.equal(getModelTargetFormat("gh", "gemini-3.7-flash"), null);
  assert.equal(getModelTargetFormat("gh", "kimi-k2.7-code"), null);
  assert.equal(ids.includes("gpt-4"), false);
  assert.equal(ids.includes("gpt-4o"), false);
--- a/tests/unit/provider-models-route.test.ts
+++ b/tests/unit/provider-models-route.test.ts
@@ -982,13 +982,6 @@ test("provider models route retries Antigravity discovery endpoints before retur
    { id: "gemini-3.7-flash-high", name: "Gemini 3.7 Flash (High)" },
    { id: "gemini-3.7-flash-medium", name: "Gemini 3.7 Flash (Medium)" },
    { id: "gemini-3.8-flash-high", name: "Gemini 3.8 Flash High" },
-    { id: "gemini-3.6-flash-high", name: "Gemini 3.6 Flash (High)" },
-    { id: "gemini-3.6-flash-medium", name: "Gemini 3.6 Flash (Medium)" },
-    { id: "gemini-3.6-flash-low", name: "Gemini 3.6 Flash (Low)" },
-    { id: "gemini-3.5-flash-extra-low", name: "Gemini 3.5 Flash (Low)" },
-    { id: "gemini-3.5-flash-low", name: "Gemini 3.5 Flash (Medium)" },
-    { id: "gemini-3-flash-agent", name: "Gemini 3.5 Flash (High)" },
-    { id: "gemini-2.5-flash", name: "Gemini 2.5 Flash" },
  ]);
 });

--- a/tests/unit/provider-registry-github-copilot-targetformat.test.ts
+++ b/tests/unit/provider-registry-github-copilot-targetformat.test.ts
@@ -40,7 +40,7 @@ const MUST_NOT_BE_RESPONSES = [
  "claude-sonnet-4.5",
  "claude-haiku-4.5",
  "gemini-3.1-pro-preview",
-  "gemini-3.5-flash",
+  "gemini-3.7-flash",
 ];

 for (const id of MUST_NOT_BE_RESPONSES) {
--- a/tests/unit/provider-registry-openai-gemini-expanded-models.test.ts
+++ b/tests/unit/provider-registry-openai-gemini-expanded-models.test.ts
@@ -13,24 +13,12 @@
 import test from "node:test";
 import assert from "node:assert/strict";

-const { openaiProvider } = await import(
-  "../../open-sse/config/providers/registry/openai/index.ts"
-);
-const { geminiProvider } = await import(
-  "../../open-sse/config/providers/registry/gemini/index.ts"
-);
+const { openaiProvider } = await import("../../open-sse/config/providers/registry/openai/index.ts");
+const { geminiProvider } = await import("../../open-sse/config/providers/registry/gemini/index.ts");

-const OPENAI_ADDED_IDS = [
-  "gpt-4.1-mini",
-  "gpt-4.1-nano",
-  "o3-mini",
-  "o4-mini",
-] as const;
+const OPENAI_ADDED_IDS = ["gpt-4.1-mini", "gpt-4.1-nano", "o3-mini", "o4-mini"] as const;

-const GEMINI_ADDED_IDS = [
-  "gemini-3.1-flash-lite",
-  "gemini-2.5-flash-lite",
-] as const;
+const GEMINI_ADDED_IDS = ["gemini-3.1-flash-lite", "gemini-2.5-flash-lite"] as const;

 test("openai registry exposes gpt-4.1 mini/nano and o3-mini/o4-mini reasoning variants", () => {
  const ids = new Set(openaiProvider.models.map((m) => m.id));
@@ -69,7 +57,7 @@ test("port did not regress previously curated openai/gemini ids", () => {
    "gemini-2.5-pro",
    "gemini-2.5-flash",
    "gemini-3.1-pro-preview",
-    "gemini-3.5-flash",
+    "gemini-3.7-flash",
  ] as const) {
    assert.ok(geminiIds.has(id), `existing gemini model ${id} must remain`);
  }
--- a/tests/unit/provider-tinycms-web.test.ts
+++ b/tests/unit/provider-tinycms-web.test.ts
@@ -74,15 +74,33 @@ test("tinycms-web is present in the provider REGISTRY with correct shape", () =>
 test("tinycms-web registry has all expected models", () => {
  const r = REGISTRY["tinycms-web"];
  assert.ok(r.models && r.models.length > 0, "must have at least one model");
-  const ids = r.models.map((m) => m.id);
-
-  assert.ok(ids.includes("gpt-5-free"), "gpt-5-free must be registered");
-  assert.ok(ids.includes("gpt-5.3-free"), "gpt-5.3-free must be registered");
-  assert.ok(ids.includes("gpt-5.3-thinking-free"), "gpt-5.3-thinking-free must be registered");
-  assert.ok(ids.includes("deepseek-v4-flash"), "deepseek-v4-flash must be registered");
-  assert.ok(ids.includes("claude-sonnet-5"), "claude-sonnet-5 must be registered");
-  assert.ok(ids.includes("gemini-3.5-flash"), "gemini-3.5-flash must be registered");
-  assert.equal(r.models.length, 16, "must have exactly 16 models");
+  assert.deepEqual(
+    r.models.map((m) => m.id),
+    [
+      "claude-fable-5",
+      "claude-opus-5",
+      "claude-sonnet-5",
+      "gpt-5.6-sol",
+      "gpt-5.6-luna",
+      "gpt-5.5",
+      "gpt-5.4-mini",
+      "gpt-5.4-nano",
+      "gpt-5.3-thinking-free",
+      "gpt-5.3-free",
+      "gpt-oss-120b",
+      "gemini-3.6-flash",
+      "gemini-3.1-pro-preview",
+      "gemini-3.1-flash-lite-preview",
+      "grok-4.5",
+      "deepseek-v4-pro",
+      "deepseek-v4-flash",
+      "kimi-k3",
+      "glm-5.2",
+      "qwen3.6-plus",
+      "mimo-v2.5-pro",
+      "mimo-v2.5",
+    ]
+  );
 });

 test("tinycms-web model names are human-readable strings", () => {
--- a/tests/unit/repro-8841-context-overflow-opencode.test.ts
+++ b/tests/unit/repro-8841-context-overflow-opencode.test.ts
@@ -60,7 +60,7 @@ function upstreamContextOverflowResponse() {
      error: {
        code: "context_length_exceeded",
        message:
-          "Input exceeds the context window for opencode/north-mini-code-free: estimated 210724 input tokens, limit 200000. Reduce the prompt or route to a model with a larger context window.",
+          "Input exceeds the context window for opencode/mimo-v2.5-free: estimated 210724 input tokens, limit 200000. Reduce the prompt or route to a model with a larger context window.",
      },
    }),
    {
@@ -71,8 +71,8 @@ function upstreamContextOverflowResponse() {
 }

 test("#8841 advertised vs compat-filter limit agree", () => {
-  const advertised = getTokenLimit("opencode-zen", "north-mini-code-free");
-  const caps = getResolvedModelCapabilities("opencode/north-mini-code-free");
+  const advertised = getTokenLimit("opencode-zen", "mimo-v2.5-free");
+  const caps = getResolvedModelCapabilities("opencode/mimo-v2.5-free");
  assert.ok(advertised > 0);
  assert.ok(
    caps.contextWindow != null && caps.contextWindow > 0,
@@ -83,7 +83,7 @@ test("#8841 advertised vs compat-filter limit agree", () => {
 test("#8841 oversized request rejected up front (no dispatch)", async () => {
  const body = largeBody();
  const pool = [
-    target("opencode/north-mini-code-free"),
+    target("opencode/mimo-v2.5-free"),
    target("opencode/hy3-free"),
  ];

@@ -96,7 +96,7 @@ test("#8841 oversized request rejected up front (no dispatch)", async () => {
      name: "pro-coding-repro-8841",
      strategy: "priority",
      models: [
-        "opencode/north-mini-code-free",
+        "opencode/mimo-v2.5-free",
        "opencode/hy3-free",
      ],
    },
--- a/tests/unit/review-reviews-v3814-fixes.test.ts
+++ b/tests/unit/review-reviews-v3814-fixes.test.ts
@@ -133,18 +133,22 @@ test("LEDGER-2: valid custom headers still pass", () => {
  assert.equal(res.success, true, res.success ? "" : JSON.stringify(res.error?.issues));
 });

-// ── LEDGER-4: every minimax-m3 registry entry is flagged multimodal ──
-test("LEDGER-4: all minimax-m3 registry entries set supportsVision (matches lite.ts)", () => {
-  const entries: { id: string; supportsVision?: boolean }[] = [];
-  for (const provider of Object.values(
+// ── LEDGER-4: MiniMax M3 vision support is provider-specific ──
+test("LEDGER-4: minimax-m3 vision metadata matches each provider", () => {
+  const entries: { provider: string; id: string; supportsVision?: boolean }[] = [];
+  for (const [provider, entry] of Object.entries(
    REGISTRY as Record<string, { models?: { id: string; supportsVision?: boolean }[] }>
  )) {
-    for (const m of provider.models || []) {
-      if (/minimax-m3/i.test(m.id)) entries.push(m);
+    for (const model of entry.models || []) {
+      if (/minimax-m3/i.test(model.id)) entries.push({ provider, ...model });
    }
  }
  assert.ok(entries.length >= 6, `expected several minimax-m3 entries, got ${entries.length}`);
-  const unflagged = entries.filter((m) => m.supportsVision !== true).map((m) => m.id);
+  const promptql = entries.find((entry) => entry.provider === "promptql");
+  assert.notEqual(promptql?.supportsVision, true, "PromptQL MiniMax M3 is text-only");
+  const unflagged = entries
+    .filter((entry) => entry.provider !== "promptql" && entry.supportsVision !== true)
+    .map((entry) => `${entry.provider}/${entry.id}`);
  assert.deepEqual(
    unflagged,
    [],
--- a/tests/unit/services/geminiRateLimitTracker.test.ts
+++ b/tests/unit/services/geminiRateLimitTracker.test.ts
@@ -47,11 +47,6 @@ test("getModelRpd strips gemma- prefix correctly for gemma models", () => {
  assert.equal(getModelRpd("gemini/gemma-4-31b-it"), 14400);
 });

-test("getModelRpd handles image-generation models (no RPM value, -1)", () => {
-  // RPD is 25 for imagen models; RPM is -1 in the JSON
-  assert.equal(getModelRpd("imagen-4-generate"), 25);
-});
-
 test("getModelRpd handles models with unlimited RPD (-1)", () => {
  // gemini-3.5-live-translate has rpd: -1
  assert.equal(getModelRpd("gemini-3.5-live-translate"), 0);
@@ -207,8 +202,8 @@ test("getModelRpm returns 0 for empty string", () => {
  assert.equal(getModelRpm(""), 0);
 });

-test("getModelRpm returns 0 for models with RPM=-1 (imagen)", () => {
-  assert.equal(getModelRpm("imagen-4-generate"), 0);
+test("getModelRpm returns 0 for models with RPM=-1", () => {
+  assert.equal(getModelRpm("gemini-3.5-live-translate"), 0);
 });

 test("getModelRpm returns 0 for null input", () => {
@@ -288,9 +283,9 @@ test("isRpmExhausted returns false for unknown model", () => {
  assert.equal(isRpmExhausted("gemini/unknown-model"), false);
 });

-test("isRpmExhausted returns false for model with RPM=-1 (imagen)", () => {
-  incrementRequestCount("imagen-4-generate");
-  assert.equal(isRpmExhausted("imagen-4-generate"), false);
+test("isRpmExhausted returns false for model with RPM=-1", () => {
+  incrementRequestCount("gemini-3.5-live-translate");
+  assert.equal(isRpmExhausted("gemini-3.5-live-translate"), false);
 });

 test("isRpmExhausted works with gemini/ prefix", () => {
--- a/tests/unit/t28-model-catalog-updates.test.ts
+++ b/tests/unit/t28-model-catalog-updates.test.ts
@@ -14,10 +14,10 @@ test("T28: gemini AI Studio catalog includes current preview models", () => {
  assert.ok(geminiIds.includes("gemini-3.1-pro-preview"));
  assert.ok(geminiIds.includes("gemini-3-flash-preview"));
  assert.ok(geminiIds.includes("gemini-3.1-flash-lite"));
-  assert.ok(geminiIds.includes("gemini-3.5-flash"));
+  assert.ok(geminiIds.includes("gemini-3.7-flash"));
  assert.ok(geminiIds.includes("gemini-2.5-flash"));
  assert.ok(geminiIds.includes("gemini-2.5-pro"));
-  assert.equal(geminiIds[0], "gemini-3.1-pro-preview", "preserve the existing Gemini default");
+  assert.equal(geminiIds[0], "gemini-3.7-flash", "use the current Gemini Flash default");
 });

 test("T28: antigravity static catalog exposes only callable Gemini tier IDs", () => {
@@ -26,12 +26,13 @@ test("T28: antigravity static catalog exposes only callable Gemini tier IDs", ()
  assert.ok(!staticIds.includes("gemini-3-pro-preview"));
  assert.ok(staticIds.includes("gemini-3.7-flash-high"));
  assert.ok(staticIds.includes("gemini-3.7-flash-medium"));
-  assert.ok(staticIds.includes("gemini-3.6-flash-high"));
-  assert.ok(staticIds.includes("gemini-3.6-flash-medium"));
-  assert.ok(staticIds.includes("gemini-3.6-flash-low"));
-  assert.ok(staticIds.includes("gemini-3.5-flash-extra-low"));
-  assert.ok(staticIds.includes("gemini-3.5-flash-low"));
-  assert.ok(staticIds.includes("gemini-3-flash-agent"));
+  assert.ok(staticIds.includes("gemini-3.7-flash-low"));
+  assert.ok(!staticIds.includes("gemini-3.6-flash-high"));
+  assert.ok(!staticIds.includes("gemini-3.6-flash-medium"));
+  assert.ok(!staticIds.includes("gemini-3.6-flash-low"));
+  assert.ok(!staticIds.includes("gemini-3.5-flash-extra-low"));
+  assert.ok(!staticIds.includes("gemini-3.5-flash-low"));
+  assert.ok(!staticIds.includes("gemini-3-flash-agent"));
  assert.ok(!staticIds.includes("gemini-3.5-flash-medium"));
  assert.ok(!staticIds.includes("gemini-3.5-flash-high"));
  assert.ok(staticIds.includes("gemini-3.1-pro-low"));
@@ -45,21 +46,21 @@ test("T28: antigravity static catalog exposes only callable Gemini tier IDs", ()
  assert.ok(!staticIds.includes("gemini-claude-opus-4-5-thinking"));
 });

-test("T28: agy free-model metadata labels upstream Gemini 3.6 tier IDs", () => {
+test("T28: agy free-model metadata labels upstream Gemini 3.7 tier IDs", () => {
  const flashNames = Object.fromEntries(
    FREE_MODEL_BUDGETS.filter(
      (entry) =>
        entry.provider === "agy" &&
-        ["gemini-3.6-flash-low", "gemini-3.6-flash-medium", "gemini-3.6-flash-high"].includes(
+        ["gemini-3.7-flash-low", "gemini-3.7-flash-medium", "gemini-3.7-flash-high"].includes(
          entry.modelId
        )
    ).map((entry) => [entry.modelId, entry.displayName])
  );

  assert.deepEqual(flashNames, {
-    "gemini-3.6-flash-low": "Gemini 3.6 Flash (Low)",
-    "gemini-3.6-flash-medium": "Gemini 3.6 Flash (Medium)",
-    "gemini-3.6-flash-high": "Gemini 3.6 Flash (High)",
+    "gemini-3.7-flash-low": "Gemini 3.7 Flash (Low)",
+    "gemini-3.7-flash-medium": "Gemini 3.7 Flash (Medium)",
+    "gemini-3.7-flash-high": "Gemini 3.7 Flash (High)",
  });
 });

@@ -90,7 +91,7 @@ test("T28: lmarena registry seeds Direct-chat Text/search; image models in IMAGE
  // Chat registry: Text + Search only (not Image thrash)
  assert.ok(lmarenaIds.length >= 40 && lmarenaIds.length < 60);
  assert.ok(lmarenaIds.includes("gemini-3.1-pro-preview"));
-  assert.ok(lmarenaIds.includes("gemini-3.5-flash-high"));
+  assert.ok(lmarenaIds.includes("gemini-3.6-flash"));
  assert.ok(lmarenaIds.includes("claude-sonnet-5"));
  assert.ok(!lmarenaIds.includes("flux-2-pro"), "image models must not live in chat registry");

--- a/tests/unit/t31-t33-t34-t38-model-specs.test.ts
+++ b/tests/unit/t31-t33-t34-t38-model-specs.test.ts
@@ -21,7 +21,7 @@ test("T31: antigravity static catalog exposes client-visible Gemini preview IDs"
  // catalog) retired the `gemini-3-pro-preview` alias, so assert the current
  // client-visible top flash tier instead.
  const staticIds = (getStaticModelsForProvider("antigravity") || []).map((m) => m.id);
-  assert.ok(staticIds.includes("gemini-3.6-flash-high"));
+  assert.ok(staticIds.includes("gemini-3.7-flash-high"));
  assert.ok(!staticIds.includes("gemini-3-pro-preview"));
  // #3303 (agy parity, discussion #3184): the Gemini + Claude budget tiers ARE
  // client-visible on the Antigravity OAuth backend (Claude was never removed).
--- a/tests/unit/upstream-retry-hints-toggle.test.ts
+++ b/tests/unit/upstream-retry-hints-toggle.test.ts
@@ -23,7 +23,7 @@ test("checkFallbackError ignores body reset text when upstream retry hints are d
    429,
    ANTIGRAVITY_RESET_TEXT,
    0,
-    "gemini-3-flash-agent",
+    "gemini-3.7-flash-high",
    "antigravity",
    null,
    antigravityProfile(false)
@@ -40,7 +40,7 @@ test("checkFallbackError honors body reset text when upstream retry hints are en
    429,
    ANTIGRAVITY_RESET_TEXT,
    0,
-    "gemini-3-flash-agent",
+    "gemini-3.7-flash-high",
    "antigravity",
    null,
    antigravityProfile(true)
--- a/tests/unit/usage-service-hardening.test.ts
+++ b/tests/unit/usage-service-hardening.test.ts
@@ -259,7 +259,7 @@ test("usage service prefers Antigravity retrieveUserQuota over catalog quotaInfo
      return new Response(
        JSON.stringify({
          models: {
-            "gemini-3-flash-agent": {
+            "gemini-3.7-flash-high": {
              quotaInfo: {
                remainingFraction: 1,
                resetTime: new Date(Date.now() + 60_000).toISOString(),
@@ -276,7 +276,7 @@ test("usage service prefers Antigravity retrieveUserQuota over catalog quotaInfo
        JSON.stringify({
          buckets: [
            {
-              modelId: "gemini-3-flash-agent",
+              modelId: "gemini-3.7-flash-high",
              remainingFraction: 0.25,
              resetTime: new Date(Date.now() + 60_000).toISOString(),
            },
@@ -294,9 +294,9 @@ test("usage service prefers Antigravity retrieveUserQuota over catalog quotaInfo
    accessToken: `ag-token-live-quota-${Date.now()}`,
  });

-  assert.equal(usage.quotas["gemini-3-flash-agent"].remainingPercentage, 25);
-  assert.equal(usage.quotas["gemini-3-flash-agent"].used, 750);
-  assert.equal(usage.quotas["gemini-3-flash-agent"].quotaSource, "retrieveUserQuota");
+  assert.equal(usage.quotas["gemini-3.7-flash-high"].remainingPercentage, 25);
+  assert.equal(usage.quotas["gemini-3.7-flash-high"].used, 750);
+  assert.equal(usage.quotas["gemini-3.7-flash-high"].quotaSource, "retrieveUserQuota");
 });

 test("usage service preserves Antigravity upstream quota bucket ids", async () => {
@@ -317,10 +317,10 @@ test("usage service preserves Antigravity upstream quota bucket ids", async () =
      return new Response(
        JSON.stringify({
          models: {
-            "gemini-3.5-flash-low": { quotaInfo: { remainingFraction: 1 } },
-            "gemini-3.5-flash-high": { quotaInfo: { remainingFraction: 1 } },
+            "gemini-3.7-flash-low": { quotaInfo: { remainingFraction: 1 } },
+            "gemini-3.7-flash-medium": { quotaInfo: { remainingFraction: 1 } },
+            "gemini-3.7-flash-high": { quotaInfo: { remainingFraction: 1 } },
            "gemini-3-flash-agent": { quotaInfo: { remainingFraction: 1 } },
-            "gemini-3.5-flash-extra-low": { quotaInfo: { remainingFraction: 1 } },
          },
        }),
        { status: 200 }
@@ -331,8 +331,8 @@ test("usage service preserves Antigravity upstream quota bucket ids", async () =
      return new Response(
        JSON.stringify({
          buckets: [
-            { modelId: "gemini-3-flash-agent", remainingFraction: 0.5 },
-            { modelId: "gemini-3.5-flash-extra-low", remainingFraction: 0.25 },
+            { modelId: "gemini-3.7-flash-high", remainingFraction: 0.5 },
+            { modelId: "gemini-3.7-flash-low", remainingFraction: 0.25 },
          ],
        }),
        { status: 200 }
@@ -347,11 +347,10 @@ test("usage service preserves Antigravity upstream quota bucket ids", async () =
    accessToken: `ag-token-legacy-buckets-${Date.now()}`,
  });

-  assert.equal(usage.quotas["gemini-3-flash-agent"].remainingPercentage, 50);
-  assert.equal(usage.quotas["gemini-3.5-flash-extra-low"].remainingPercentage, 25);
-  assert.equal(usage.quotas["gemini-3.5-flash-low"].remainingPercentage, 100);
-  assert.equal(usage.quotas["gemini-3.5-flash-medium"], undefined);
-  assert.equal(usage.quotas["gemini-3.5-flash-high"], undefined);
+  assert.equal(usage.quotas["gemini-3.7-flash-high"].remainingPercentage, 50);
+  assert.equal(usage.quotas["gemini-3.7-flash-low"].remainingPercentage, 25);
+  assert.equal(usage.quotas["gemini-3.7-flash-medium"].remainingPercentage, 100);
+  assert.equal(usage.quotas["gemini-3-flash-agent"], undefined);
 });

 test("usage service retries Antigravity fetchAvailableModels across the shared fallback order", async () => {
--- a/tests/unit/veoaifree-web-executor.test.ts
+++ b/tests/unit/veoaifree-web-executor.test.ts
@@ -52,7 +52,6 @@ test("detectIntent returns 'video' for veo models", () => {
 test("detectIntent returns 'image' for image models", () => {
  assert.equal(detectIntent("image-gen"), "image");
  assert.equal(detectIntent("banana"), "image");
-  assert.equal(detectIntent("imagen-4"), "image");
  assert.equal(detectIntent("nano-banana"), "image");
 });