fix(model): return clear error instead of silent openai default for unrecognized models (#2492)

Integrated into release/v3.8.2

open-sse/services/model.ts

@@ -468,11 +468,17 @@ async function resolveModelByProviderInference(modelId: string, extendedContext:
     return { provider: "gemini", model: modelId, extendedContext };
   }
 
-  // Last resort: treat as openai model
+  // Last resort: no provider could be inferred — return a clear error instead
+  // of silently defaulting to "openai", which would produce a misleading
+  // "No credentials for provider: openai" response when the model name
+  // is unrecognised (e.g. a missing combo, a typo, or a bare model id
+  // that doesn't exist in any provider's catalog).
   return {
-    provider: "openai",
+    provider: null,
     model: modelId,
     extendedContext,
+    errorType: "model_not_found",
+    errorMessage: `Unable to determine provider for model '${modelId}'. Use a provider/model prefix (e.g. openai/${modelId}) or ensure the model is added as a combo entry.`,
   };
 }
 

src/sse/handlers/chatHelpers.ts

@@ -175,6 +175,17 @@ export async function resolveModelOrError(
   }
 
   if (!modelInfo.provider) {
+    // model_not_found: raised by resolveModelByProviderInference when no
+    // provider could be inferred — return a clear error instead of the
+    // misleading "openai" default that the old code silently fell back to.
+    if ((modelInfo as any).errorType === "model_not_found") {
+      const message =
+        (modelInfo as any).errorMessage ||
+        `Model '${modelStr}' could not be resolved to a known provider.`;
+      log.warn("CHAT", message, { model: modelStr });
+      return { error: errorResponse(HTTP_STATUS.BAD_REQUEST, message) };
+    }
+
     if ((modelInfo as any).errorType === "ambiguous_model") {
       // Family disambiguation: if the model name begins with a known
       // non-OAuth family prefix, auto-pick the family-native provider

tests/unit/chat-helpers.test.ts

@@ -387,3 +387,17 @@ test("withSessionHeader adds headers to mutable and immutable responses", async
   assert.equal(immutable.status, 302);
   assert.equal(await immutable.text(), "");
 });
+
+test("resolveModelOrError returns model_not_found error for unrecognised bare model names", async () => {
+  const result = await resolveModelOrError(
+    "completely-unknown-model-xyz",
+    { messages: [{ role: "user", content: "hello" }] },
+    "/v1/chat/completions"
+  );
+
+  assert.ok(result.error);
+  assert.equal(result.error.status, 400);
+  const json = (await result.error.json()) as any;
+  assert.match(json.error.message, /Unable to determine provider/i);
+  assert.match(json.error.message, /completely-unknown-model-xyz/i);
+});

tests/unit/services-branch-hardening.test.ts

@@ -354,11 +354,11 @@ test("model helpers cover malformed input, alias maps, wildcard aliases, ambigui
     model: "gemini-custom",
     extendedContext: false,
   });
-  assert.deepEqual(await modelService.getModelInfoCore("made-up-model", {}), {
-    provider: "openai",
-    model: "made-up-model",
-    extendedContext: false,
-  });
+  const unknownModel = await modelService.getModelInfoCore("made-up-model", {});
+  assert.equal(unknownModel.provider, null);
+  assert.equal(unknownModel.errorType, "model_not_found");
+  assert.ok(unknownModel.errorMessage.includes("made-up-model"));
+  assert.ok(unknownModel.errorMessage.includes("provider/model prefix"));
 });
 
 test("error classifier covers empty-content helpers, context overflow and remaining error classes", () => {