Troubleshooting errors
Every API and Console error you can hit, what it means and how to fix it.
- How do I fix 400 invalid_json or invalid_request on the Model API?The body isn't valid JSON or isn't a valid chat completion. Check quoting, Content-Type and the messages array.Troubleshooting
- How do I fix 400 unsupported_content or unsupported_parameter?The Model API is text-only and doesn't support n > 1 or logprobs. Remove image parts and unsupported fields.Troubleshooting
- How do I fix 401 account_mismatch?The Account ID you sent doesn't own the key. Send the Account ID of the account that created the key.Troubleshooting
- How do I fix 401 expired_api_key?The key reached the expiry date chosen when it was created. Create a new key and replace it.Troubleshooting
- How do I fix 401 invalid_api_key?The key is wrong, revoked or cut off. Check the full clsk_ key, that it's Active in the Console, and the Bearer format.Troubleshooting
- How do I fix 401 missing_account_id?Your request has no X-Clatter-Account header. Add it with your CL- Account ID next to the API key.Troubleshooting
- How do I fix 403 insufficient_scope?The key lacks the permission this endpoint needs. Create a key with that permission, such as models:invoke or runs:write.Troubleshooting
- How do I fix 403 team_scope_denied?You're on a team and the owner didn't give you access to this product. Ask the owner to add it on the Team page.Troubleshooting
- How do I fix 404 model_not_found on the Model API?The model id isn't one Clatter Labs serves. Use supercosmos-flash, supercosmos-max, supercosmos-heavy or supercosmos with an effort.Troubleshooting
- How do I fix 410 model_retired?You called a retired model id, such as a Line Code id. Switch to supercosmos-flash, -max or -heavy. Not billed.Troubleshooting
- How do I fix 413 too_large or request_too_large?The request is over the size limits: Agent API input 20,000 characters / 64 KB body; Model API 3 MB or 800 messages.Troubleshooting
- How do I fix 429 rate_limited?The key made more requests per minute than its limit. Wait the Retry-After seconds, back off, or use a higher-limit key.Troubleshooting
- List of Clatter Labs API error codesEvery error code by HTTP status for the Agent API and Model API, with a link to what to do.Reference
- My Model API stream ended with an error event. What happened?A stream that fails midway sends data: {"error": …} then data: [DONE]. Retry the call; check status if it repeats.Troubleshooting
- What do Clatter Labs API errors look like?Errors are JSON with a stable code and a message. The Model API uses the OpenAI error shape and returns X-Request-Id.Reference
- What does 403 account_suspended mean, and how long does it last?Your account is paused for 24 hours under the abuse policy. Wait until resets_at, or appeal through Support.Troubleshooting
- What is the x-clatter-notice: abuse_warning header?A first strike under the abuse policy. The request still succeeds, with x-clatter-notice and the warning text.FAQ
- What should I do about model_overloaded, model_unavailable or model_timeout?The model is at capacity or briefly unavailable. Retry with backoff, honouring Retry-After when it's sent.Troubleshooting
- What should I include when I contact support about an API error?Your Account ID, the error code and message, the time with time zone, the endpoint, the X-Request-Id or run id. Never the key.How-to
Still need help?
Tried the steps above and still stuck? Email us. We usually reply within the first hour, and always within 24 hours. A full support team is with you.
Infrastructure questions, or permanent deletion of your data
infrastructure@clatterlabs.xyzFor deleting your whole account or the data of one service.