Errors - ModelsLab

The Enterprise LLM Endpoint uses standard HTTP status codes, so OpenAI SDKs raise their normal exceptions, for example AuthenticationError for 401.

Status codes

Status	Meaning	What to do
`400`	The model rejected the request body, for example an unsupported parameter.	Read `error.message`, then correct the request.
`401`	The API key is missing, wrong, or belongs to a different enterprise server.	Send your enterprise API key as `Authorization: Bearer <key>`.
`402`	The enterprise plan is stopped because of a payment problem.	Update your payment in the ModelsLab dashboard.
`403`	The plan is not active: the trial expired, or the plan was cancelled or paused.	Buy or renew your enterprise plan, or contact support.
`404`	The path is not one of the endpoints that your plan allows.	Use `chat/completions`, `completions`, or `models`.
`413`	The request is larger than 30 MB.	Send large files as a URL instead of base64. See Files, Images and PDFs.
`429`	Too many requests reached the model at the same time.	Wait, then try again with backoff.
`502`	The model did not respond in time.	Try again. For long outputs, set `"stream": true`.
`503`	The endpoint is not set up yet.	Contact support.
`5xx`	The model had an error.	Try again with backoff.

Error body

Errors from the endpoint and from the model use the OpenAI error format:

json

{
  "error": {
    "message": "This enterprise plan is not active",
    "type": "modelslab_proxy_error",
    "code": 403
  }
}

Errors about the API key (401), payment (402), and an expired trial (403) use a short ModelsLab format:

json

{
  "status": "error",
  "message": "Invalid enterprise API key for proxy server"
}