Errors

Meto AI relay errors generally follow OpenAI-style JSON bodies:

{
  "error": {
    "message": "Human-readable description",
    "type": "invalid_request_error",
    "code": "optional_code"
  }
}

Common HTTP status codes

Status	Typical cause
400	Invalid request body or unsupported parameter
401	Missing or invalid API key
403	Model not allowed, insufficient permissions, or policy block
404	Unknown route or model
429	Rate limit exceeded — retry with backoff
500	Upstream or platform error — retry later
503	Temporary overload or maintenance

Rate limits

Rate limits may apply per key, account, or model. When limited:

1. Read the Retry-After header when present
2. Exponential backoff on 429 responses
3. Contact support if you need higher sustained throughput

Getting help

Include the following when contacting support@metotoken.ai:

• Approximate timestamp (UTC)
• Model name and endpoint
• Request ID from response headers if available
• Redacted request/response samples (never send live API keys)