diff --git a/changelog.d/fixes/8307-codex-image-account-fallback-retryable.md b/changelog.d/fixes/8307-codex-image-account-fallback-retryable.md
new file mode 100644
index 0000000000..bd4a70a1ab
--- /dev/null
+++ b/changelog.d/fixes/8307-codex-image-account-fallback-retryable.md
@@ -0,0 +1 @@
+- **fix(images):** retry Codex image generation on a sibling ChatGPT account when the requested model isn't entitled on the current account, instead of failing the request outright ([#8307](https://github.com/diegosouzapw/OmniRoute/pull/8307)).
diff --git a/open-sse/handlers/imageGeneration.ts b/open-sse/handlers/imageGeneration.ts
index 7fb6357db1..a8452f0245 100644
--- a/open-sse/handlers/imageGeneration.ts
+++ b/open-sse/handlers/imageGeneration.ts
@@ -184,6 +184,29 @@ function sanitizeImageProviderError(errorText: string): unknown {
   return sanitizeErrorMessage(errorText);
 }
 
+// #8307 — some ChatGPT accounts can run Codex but lack entitlement for the specific
+// requested image model. Upstream signals this as a 400 with an exact, stable message
+// (not a generic "invalid request"). Classify it so the caller can mark the failure
+// `retryable: true`, which routes it through the same sibling-account fallback that
+// already handles 401s (executeImageWithCredentialFallback, src/sse/services/imageCredentialRetry.ts).
+function isCodexChatGptModelAccessError(status: number, errorText: string, model: string): boolean {
+  if (status !== 400) return false;
+  const parsed = parseJsonOrNull(errorText);
+  let detail: string | null = null;
+  if (typeof parsed === "string") {
+    detail = parsed;
+  } else if (parsed && typeof parsed === "object") {
+    const obj = parsed as Record<string, unknown>;
+    if (typeof obj.detail === "string") detail = obj.detail;
+    else if (typeof obj.message === "string") detail = obj.message;
+    else if (obj.error && typeof obj.error === "object") {
+      const nested = (obj.error as Record<string, unknown>).message;
+      if (typeof nested === "string") detail = nested;
+    }
+  }
+  return detail === `The '${model}' model is not supported when using Codex with a ChatGPT account.`;
+}
+
 const BFL_MODEL_ENDPOINTS = {
   "flux-2-max": "/v1/flux-2-max",
   "flux-2-pro": "/v1/flux-2-pro",
@@ -2532,6 +2555,7 @@ async function handleCodexImageGeneration({
       const safeErrorLog =
         typeof safeError === "string" ? safeError : JSON.stringify(safeError ?? {});
       if (log) log.error("IMAGE", `${provider} error ${response.status}: ${safeErrorLog}`);
+      const retryable = isCodexChatGptModelAccessError(response.status, errorText, model);
       return {
         ok: false as const,
         error: {
@@ -2542,6 +2566,7 @@ async function handleCodexImageGeneration({
           error: safeError,
           requestBody: requestBodyForLog,
           path: logPath,
+          ...(retryable ? { retryable: true } : {}),
         },
       };
     }
diff --git a/tests/unit/image-generation-handler.test.ts b/tests/unit/image-generation-handler.test.ts
index d525e60d4d..49709a13c3 100644
--- a/tests/unit/image-generation-handler.test.ts
+++ b/tests/unit/image-generation-handler.test.ts
@@ -2026,3 +2026,57 @@ test("handleImageGeneration (codex) forwards size and maps GPT-Image quality to
     globalThis.fetch = originalFetch;
   }
 });
+
+// #8307 — some ChatGPT accounts can run Codex but lack entitlement for the specific
+// requested image model, and the upstream 400 for that exact case is retryable on a
+// sibling account: executeImageWithCredentialFallback (route.ts) already retries on
+// this signal when the handler marks the failure `retryable: true` — mirroring the
+// existing 401 auto-rotate path, no new retry loop needed in the handler itself.
+test("handleImageGeneration (codex) marks the ChatGPT-account model-access 400 as retryable", async () => {
+  const originalFetch = globalThis.fetch;
+  globalThis.fetch = async () =>
+    new Response(
+      JSON.stringify({
+        error: {
+          message:
+            "The 'gpt-5.6-sol' model is not supported when using Codex with a ChatGPT account.",
+        },
+      }),
+      { status: 400, headers: { "content-type": "application/json" } }
+    );
+
+  try {
+    const result = await handleImageGeneration({
+      body: { model: "codex/gpt-5.6-sol", prompt: "kitten" },
+      credentials: { accessToken: "codex-token" },
+      log: null,
+    });
+    assert.equal(result.success, false);
+    assert.equal(result.status, 400);
+    assert.equal(result.retryable, true);
+  } finally {
+    globalThis.fetch = originalFetch;
+  }
+});
+
+test("handleImageGeneration (codex) does not mark an ordinary 400 as retryable", async () => {
+  const originalFetch = globalThis.fetch;
+  globalThis.fetch = async () =>
+    new Response(JSON.stringify({ error: { message: "Invalid prompt" } }), {
+      status: 400,
+      headers: { "content-type": "application/json" },
+    });
+
+  try {
+    const result = await handleImageGeneration({
+      body: { model: "codex/gpt-5.6-sol", prompt: "kitten" },
+      credentials: { accessToken: "codex-token" },
+      log: null,
+    });
+    assert.equal(result.success, false);
+    assert.equal(result.status, 400);
+    assert.equal(result.retryable, undefined);
+  } finally {
+    globalThis.fetch = originalFetch;
+  }
+});