generated: '2026-07-20'
method: searched
source: https://docs.mindlogic.ai/docs/general/api-gateway/reference/errors
docs: https://docs.mindlogic.ai/docs/general/api-gateway/reference/errors
description: >-
  Error reference for the Mindlogic FactChat API Gateway. Codes below are
  documented on the Gateway error-reference page; the error envelope shape was
  confirmed against a live unauthenticated request to
  https://factchat-cloud.mindlogic.ai/v1/gateway/models/.
envelopes:
  - name: gateway_error
    shape: '{"detail": {"code": <int>, "message": "<type> - <human message>"}}'
    example: '{"detail": {"code": 401, "message": "authentication_error - Bearer token is not supplied"}}'
    source: live probe (401 from /v1/gateway/models/)
  - name: validation_error
    format: fastapi
    shape: '{"detail": [{"loc": [...], "msg": "<message>", "type": "<error type>"}]}'
    schema: openapi/mindlogic-api-gateway-openapi-original.json#/components/schemas/HTTPValidationError
    note: FastAPI-style 422 validation error, referenced by most operations.
errors:
  - status: 400
    name: Bad Request
    message: Invalid parameters, or a model type that does not match the endpoint.
    remediation: Validate parameters; verify the model is compatible with the endpoint used.
  - status: 400
    name: Bad Request
    message: "Model 'X' is not an OpenAI model"
    remediation: Switch to /chat/completions/ or the Claude (Anthropic) endpoints as appropriate.
  - status: 401
    name: Unauthorized
    message: Authorization header is not supplied
    remediation: "Add an Authorization: Bearer KEY or x-api-key: KEY header."
  - status: 401
    name: Unauthorized
    message: Bearer token is not supplied
    remediation: "Use the Authorization: Bearer YOUR_KEY format."
  - status: 401
    name: Unauthorized
    message: Invalid API key
    remediation: Verify the key exists in developer settings (FactChat > API Gateway).
  - status: 402
    name: Payment Required
    message: Credit balance exhausted
    remediation: Recharge credits in developer settings.
  - status: 404
    name: Not Found
    message: "Model 'X' not found"
    remediation: Check GET /v1/gateway/models/ and verify the model is activated.
  - status: 404
    name: Not Found
    message: "Model 'X' is not an Anthropic model"
    remediation: Use /chat/completions/ for non-Anthropic models.
  - status: 404
    name: Not Found
    message: Chatbot not found
    remediation: Verify the chatbot_id path parameter references an existing chatbot.
  - status: 422
    name: Validation Error
    message: Request body/parameters failed validation.
    remediation: Inspect detail[].loc / detail[].msg for the offending field.
  - status: 429
    name: Rate Limited
    message: Rate limit exceeded
    remediation: Implement an exponential-backoff retry strategy.
  - status: 500
    name: Server Error
    message: Upstream provider error.
    remediation: Contact support with the error details.
common_issues:
  - issue: Reasoning + tools conflict
    behavior: The Gateway auto-removes reasoning_effort on retry.
  - issue: Empty reasoning-model responses
    behavior: "Requires max_tokens 16000+; some models also need temperature: 1."