generated: '2026-07-20'
method: searched
source: https://docs.withmartian.com/api-reference/errors
format: openai-compatible
envelope:
  openai: '{ "error": { "message": string, "type": string, "param": string|null, "code": string } }'
  anthropic: '{ "type": "error", "error": { "type": string, "message": string } }'
http_status_codes:
- status: 200
  name: OK
  meaning: The request was successful; the response body contains the requested data.
- status: 400
  name: Bad Request
  meaning: The request was malformed or contains invalid parameters. Check the request body and parameters.
- status: 401
  name: Unauthorized
  meaning: Authentication failed. The API key is missing, invalid, or revoked.
- status: 403
  name: Forbidden
  meaning: No permission to access the resource. Check account status and credits.
- status: 404
  name: Not Found
  meaning: The requested resource, endpoint, or model does not exist. Check the URL and model name.
- status: 422
  name: Unprocessable Entity
  meaning: The request was well-formed but has semantic errors (e.g. unsupported model parameters).
- status: 429
  name: Too Many Requests
  meaning: Rate limit hit. Implement backoff and retry logic.
- status: 500
  name: Internal Server Error
  meaning: An unexpected server error occurred. Retry or contact support if it persists.
- status: 503
  name: Service Unavailable
  meaning: Service temporarily unavailable; the upstream provider may be down. Retry later.
error_codes:
- code: invalid_api_key
  type: invalid_request_error
  status: 401
  meaning: API key missing, malformed, revoked, or expired.
  remediation: Verify the key in the Martian Dashboard and use the Authorization Bearer header.
- code: insufficient_credits
  type: insufficient_quota
  status: 403
  meaning: Account has insufficient credits.
  remediation: Add credits to the account in the Martian Dashboard.
- code: model_not_found
  type: invalid_request_error
  status: 404
  meaning: The named model does not exist or is missing the provider prefix.
  remediation: Use a provider/model-name from the Available Models page.
- code: invalid_parameter
  type: invalid_request_error
  status: 400
  meaning: A parameter is missing, wrong type, or out of range.
  remediation: Review the API Endpoints docs and correct the parameter.
- code: rate_limit_exceeded
  type: rate_limit_error
  status: 429
  meaning: Too many requests in a window.
  remediation: Back off and retry after the advised interval.
notes: >-
  Errors are OpenAI-compatible by default; requests to /v1/messages return the
  Anthropic-compatible error envelope. This is an API-level error registry, not
  an RFC 9457 problem-details catalog and not payment decline codes.