fix(onboard): restore openai-responses API for all Azure URLs

OpenClaw · OpenClaw · commit ba811b4629c0 · 2026-03-20T08:31:31.000+07:00
Regression introduced in 91104ac broke local vLLM endpoints that rely on openai-responses API compatibility. The commit narrowed Azure detection from 'isAzure' to 'isAzureOpenAi' (*.openai.azure.com only), causing non-Azure OpenAI-compatible endpoints to route to /v1/chat/completions instead of /responses. This resulted in 404 errors for local vLLM setups configured with openai-responses mode. Fix: Revert providerApi logic to use 'isAzure' (covers both *.services.ai.azure.com and *.openai.azure.com) to preserve backward compatibility. Fixes #50719
diff --git a/src/commands/onboard-custom.test.ts b/src/commands/onboard-custom.test.ts
@@ -506,7 +506,7 @@ describe("applyCustomApiConfig", () => {
     const provider = result.config.models?.providers?.[providerId];
 
     expect(provider?.baseUrl).toBe("https://my-resource.services.ai.azure.com/openai/v1");
-    expect(provider?.api).toBe("openai-completions");
+    expect(provider?.api).toBe("openai-responses");
     expect(provider?.authHeader).toBe(false);
     expect(provider?.headers).toEqual({ "api-key": "key123" });
 
diff --git a/src/commands/onboard-custom.ts b/src/commands/onboard-custom.ts
@@ -686,7 +686,7 @@ export function applyCustomApiConfig(params: ApplyCustomApiConfigParams): Custom
     normalizeOptionalProviderApiKey(params.apiKey) ??
     normalizeOptionalProviderApiKey(existingApiKey);
 
-  const providerApi = isAzureOpenAi
+  const providerApi = isAzure
     ? ("openai-responses" as const)
     : resolveProviderApi(params.compatibility);
   const azureHeaders = isAzure && normalizedApiKey ? { "api-key": normalizedApiKey } : undefined;
