Skip to main content

FAQ

Which value should I use for base_url?​

Use {{BASE_URL}} in tools that ask for an OpenAI-compatible base URL or endpoint root.

Which value should I use for the full chat completions endpoint?​

Use {{CHAT_COMPLETIONS_URL}} when a tool asks for the full request URL instead of a base URL.

My tool says the model does not exist​

Copy the model name directly from model marketplace. Use the exact request name shown in the console.

My request returns 401 or 403​

Check that the API key is active, copied completely, and sent as Authorization: Bearer <key>.

My request returns 404​

Check the URL first. Base URL fields should use {{BASE_URL}}. Full chat endpoint fields should use {{CHAT_COMPLETIONS_URL}}.

My request returns 429​

Check whether the current workload has reached the quota or the current RPM / TPM limit. Then inspect Usage and Teams.

My request reaches the gateway but still fails​

Open Logs and inspect status code, latency, model name, and the returned error message. Logs are the fastest way to separate config issues from model-side issues.

Which page should I open first?​

Use this order:

  1. Getting Started for the onboarding path
  2. API Keys for credentials
  3. model marketplace for the model name
  4. Teams to confirm the active team context
  5. Playground for the first validation if your role includes write access
  6. Logs for troubleshooting

I need embeddings instead of chat completions​

Keep the same base_url and key. Use {{EMBEDDINGS_URL}} only when a tool asks for the full embeddings endpoint.