Understand error codes - Writer AI Studio

Error Code Reference

This page provides a comprehensive reference for all error codes, HTTP status codes, and error handling patterns used across Writer APIs.

Authentication Issues

401 Unauthorized: check your API key format and permissions

Bad Request Data

400 + validation_error: verify required fields and data types

Rate Limiting

429 Too Many Requests: implement exponential backoff

Server Problems

5xx Errors: check our status page or retry your request

Error Response Format

All API endpoints return errors in a consistent JSON format:

{
  "tpe": "error_type",
  "errors": [
    {
      "description": "Human-readable error description",
      "key": "error.message.key",
      "extras": {}
    }
  ],
  "extras": {}
}

Response Structure

🔧 Core fields (always present)

Field	Type	Description
`tpe`	string	Error type identifier (for example, “BadRequest”, “Unauthorized”)
`errors`	array	Array of error messages with details
`extras`	object	Additional error context and metadata

HTTP Status Codes

Writer APIs use standard HTTP status codes to indicate the nature of errors:

Status	Error code	Type	Description	Common causes
400	`BadRequest`	Validation	Invalid request parameters or malformed request body	Missing required fields, invalid format
401	`Unauthorized`	Authentication	Missing or invalid authentication credentials	Missing API key, invalid format
403	`Forbidden`	Authorization	Valid credentials but insufficient permissions	Insufficient access rights
404	`NotFound`	Validation	Requested resource does not exist	Invalid resource ID, deleted resource
409	`Conflict`	Validation	Request conflicts with current state	Duplicate resource, version conflict
410	`Gone`	Validation	Resource was permanently removed	Deleted/expired resource
413	`PayloadTooLarge`	Validation	Request body too large	Exceeds configured size limits
422	`UnprocessableEntity`	Validation	Semantically invalid request	Fails business rules/constraints
429	`TooManyRequests`	Rate Limiting	Rate limit or quota exceeded	Too many requests per time period
500	`InternalServerError`	Server	Internal server error	Server-side processing issue
503	`ServiceUnavailable`	Server	Service temporarily unavailable	Maintenance, high load
507	`InsufficientStorage`	Server	Insufficient storage to complete request	Storage quota exhausted

Error Codes by Service

This section provides a comprehensive reference of all possible error codes organized by service and endpoint.

Chat Completions API

Endpoint: POST /v1/completion

Error Type	HTTP Status	Description	Possible Causes
`BadRequest`	400	Invalid model parameter	Model name not supported or malformed
`InvalidInputParameters`	400	Invalid prompt format	Empty prompt or invalid encoding
`InvalidContentType`	400	Invalid parameters	Invalid `temperature`, `max_tokens`, or other parameters
`BadCredentials`	401	Invalid API key	Missing or malformed API key
`BadToken`	401	API key expired	API key has expired
`InsufficientAccessRights`	403	Insufficient permissions	API key lacks required permissions
`QuotaExceeded`	429	Rate limit exceeded	Too many requests per minute
`CompletionModelQuotaError`	429	Quota exceeded	Monthly usage limit reached
`CompletionModelUnavailableError`	500	Model unavailable	AI model temporarily unavailable
`CompletionModelInternalError`	500	Service overloaded	High load, retry with backoff

Guardrail Errors

Guardrail errors occur when content is blocked by configured safety controls.

Error Type	HTTP Status	Description	Possible Causes
`GuardrailBlockedError`	400	Content blocked by guardrail	Input or output violated guardrail policy
`BadRequest`	400	Bedrock guardrail violation	Content blocked by AWS Bedrock guardrail policy
`BadRequest`	400	Azure Content Safety violation	Content blocked by Azure text moderation or Prompt Shields
`BadCredentials`	401	Guardrail authentication failed	Invalid AWS credentials for Bedrock guardrail
`InsufficientAccessRights`	403	Guardrail access denied	Missing permissions for guardrail provider
`ResourceNotFound`	404	Guardrail not found	Invalid guardrail ID or version
`InternalError`	500	Guardrail processing error	Error communicating with guardrail provider
`InternalError`	503	Guardrail service unavailable	Guardrail provider service is down

Troubleshooting Common Errors

400 Bad Request

Verify all required parameters are included
Check parameter value formats and constraints
Ensure request body is valid JSON

401 Unauthorized

Verify API key is correct and not expired
Check that the API key has the required permissions
Ensure the Authorization header is properly formatted

403 Forbidden

Check if your organization/team has access to the requested resource
Verify that the feature is enabled for your plan
Contact support if you believe this is an error

429 Too Many Requests

Implement exponential backoff retry logic
Check your current usage against rate limits
Consider upgrading your plan for higher limits