Diagnose OpenAI-Compatible API Failures.
One endpoint. Six real probes. Evidence for the network, auth, model, server, and SDK layer—before you lose another hour blaming the wrong thing.
Test the endpoint, not your assumptions.
Run six focused probes and receive evidence for each result.
Check model listing, basic chat, SSE streaming, tool calling, the tool-result round trip, and JSON mode.
- Input
- Base URL, API key, model
- Output
- Evidence, fixes, reproduction code
Example report
Illustrative report · generated with the real P0 report shape
Reads /models, confirms models are enumerable, and checks whether the entered model ID is present.
24 msSends a minimal text request and checks whether the response shape can be parsed by OpenAI SDKs.
41 msThe stream completes, but one frame uses a provider-specific field.
58 msThe endpoint accepted tools but did not return a tool_calls array.
75 msSkipped because the tool-call prerequisite did not pass.
92 msUses response_format: json_object and checks whether the endpoint reliably returns valid JSON.
109 msEnable tool calling for this model or update the compatibility server.
Troubleshooting guides
Fast answers for configuration and compatibility failures, backed by requests you can reproduce.
What this proves
A green response is not compatibility. Each probe records assertions, the failure layer, expected versus actual data, a repair path, and a shareable reproduction.
Methodology & privacy