When all combo models are exhausted (502/503), OmniRoute now checks for
a globalFallbackModel setting and attempts one last request through it
before returning the error. Settings stored in key_value table, no
migration needed.
Adds a new /dashboard/cache page that surfaces the existing but UI-less
semantic cache infrastructure.
Changes:
- New page: src/app/(dashboard)/dashboard/cache/page.tsx
- Live stats: memory entries, DB entries, cache hits, tokens saved
- Hit rate progress bar with color coding (green/yellow/red)
- Hits/Misses/Total breakdown
- Idempotency layer stats (active dedup keys + window)
- Cache behavior info panel
- Clear All button
- Auto-refresh every 10s
- Enhanced API: src/app/api/cache/route.ts
- DELETE ?model=<name> — invalidate by model
- DELETE ?signature=<hex> — invalidate single entry
- DELETE ?staleMs=<ms> — invalidate entries older than N ms
- DELETE (no params) — clear all (existing behavior)
- Sidebar: added Cache nav item (icon: cached)
- i18n: added cache + sidebar.cache keys for all 31 supported locales
No new dependencies. All functionality builds on existing semanticCache.ts,
cacheLayer.ts, and idempotencyLayer.ts modules.
Co-authored-by: oyi77 <oyi77@users.noreply.github.com>
- Add codexAuthFile.ts utility: builds Codex auth.json payload from OAuth connection
(id_token, access_token, refresh_token, account_id) with auto-refresh if expired
- Add POST /api/providers/[id]/codex-auth/export: downloads auth.json file
- Add POST /api/providers/[id]/codex-auth/apply-local: writes auth.json to local CLI path
- Add 'Apply auth' and 'Export auth' buttons to ConnectionRow (Codex provider only)
- Add i18n keys for en and pt-BR
- introduce open-sse/translator/helpers/schemaCoercion.ts to coerce
numeric JSON Schema fields encoded as strings
- wire coerceToolSchemas and sanitizeToolDescriptions into translator
pipeline; ensure tool descriptions are sanitized
- inject empty reasoning content for tool calls when target is OpenAI
format
- update qwen base URL to DashScope-compatible endpoint
- extend antigravity static catalog with Gemini 3.1 pro preview models and
update Gemini model specs with preview aliases
- implement call log max cap caching with TTL; expose invalidateCallLogsMaxCache
and invalidate on settings PATCH
- add tests: call-log-cap.test.mjs and tool-request-sanitization.test.mjs;
extend tests for Windsurf integration and gemini previews
- update CLI runtime and tools to include Windsurf as a guide-only tool
- add maxCallLogs to validation schemas (settings and updateSettings)
- add Czech README (README.cs.md) to repository
The 'Import from /models' button failed because opencode-zen was not
registered in PROVIDER_MODELS_CONFIG. The provider's API at
https://opencode.ai/zen/v1/models returns standard OpenAI-compatible
format and is now properly configured for model import.
Reviewed and approved via consolidated analysis. Turkish locale (31st language) follows existing i18n patterns perfectly. Registered in config.ts, generate-multilang.mjs, and full tr.json translation file.
Reviewed and approved via consolidated analysis. GLM-5.1 addition and pricing corrections match official Z.AI pricing page. All 5 files follow existing patterns.
- Add glm-5.1 model to GLM Coding provider with fitness scores
- Update glm-5 pricing to match Z.AI API page ($1/$3.2/$0.2)
- Set glm-5.1 pricing to $1.2/$5/$0.3 per Z.AI
- Remove glm-4-32b (deprecated, returns empty from upstream)
- Rename Z.AI provider display name from "Z.AI (GLM-5)" to "Z.AI"
- Update zai pricing section to match glm pricing
The Clear All Models button was only deleting custom models from the
database but leaving their aliases intact, so the UI didn't reflect
the deletion. Now it also deletes all aliases belonging to the provider
and refreshes the alias state.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
High backoffLevel (up to 15) persisted permanently in the DB after a burst of 429s. The account health score dropped to zero (100 - 15*10 = -50), causing the account selector to never pick the account again. Only a successful request could reset backoffLevel via clearAccountError, but the account was never selected — creating a deadlock.
Now, during account selection, any non-terminal connection whose rateLimitedUntil has passed gets its backoffLevel reset to 0 and testStatus restored to active. The DB update is fire-and-forget to avoid blocking the hot path.
Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-opencode)
Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
Thanks @brendandebeasi for this excellent contribution! 🎉 The bounded retry with exponential backoff is exactly the right approach for expired connections. Merged and will be included in v3.1.1.
Connections marked as 'expired' were permanently skipped by the health check scheduler (line 176: if testStatus === "expired" return). A single transient refresh failure could permanently disable auto-refresh, requiring manual re-authentication.
Replace the hard skip with a bounded retry mechanism: up to 3 attempts with exponential backoff (5min, 10min, 20min). On success, the connection is fully restored to active. On exhaustion, it remains expired (same as before). The existing circuit breaker (5 failures → 30min pause) provides additional protection.
Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-opencode)
Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
OpenAI-compatible clients (OpenCode, etc.) check capabilities/input_modalities fields on the /v1/models response to determine if a model supports image input. Omniroute was not emitting these fields, causing clients to assume text-only for all models routed through the proxy.
Add keyword-based vision detection (matching the existing playground heuristic) that annotates model entries with capabilities:{vision:true}, input_modalities:["text","image"], and output_modalities:["text"] for known multimodal models (GPT-4o/4-turbo, Claude 3+, Gemini, Pixtral, Qwen-VL, etc.).
Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-opencode)
Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
- Rename in.json → hi.json: 'in' is Indonesian (ISO 639-1), Hindi is 'hi'.
Fixes Weblate locale conflict where id.json and in.json both claimed Indonesian.
- Move empty tool name filter before Codex passthrough: nativeCodexPassthrough
skipped all input sanitization, causing 400 'empty tool name' from upstream.
- Collapse 3+ consecutive newlines to \n\n in response sanitizer: thinking
models accumulate excessive line breaks between tool call blocks.
- OpenAI-to-Claude translator now maps reasoning_effort (low/medium/high/max)
to Claude's thinking.budget_tokens. Fixes clients like OpenCode sending
reasoning_effort via @ai-sdk/openai-compatible losing thinking configuration.
- Ensures max_tokens > budget_tokens for all thinking configs.
- Token health check now proactively refreshes tokens within 5 min of expiry,
regardless of the configured health check interval — addresses Qwen OAuth
token refresh failures between scheduled checks.
