Error Codes

When an API request fails, the response includes an HTTP status code and an error message. Here are the common errors and how to resolve them.

HTTP Status Codes

Code	Name	Description	Resolution
400	Bad Request	Invalid request format or parameters	Check your request body and parameters
401	Unauthorized	Missing or invalid API key	Verify your Authorization: Bearer header
403	Forbidden	API key lacks permission for this resource	Check your plan tier or contact support
404	Not Found	Invalid endpoint or model not found	Verify the URL and model name
429	Too Many Requests	Rate limit exceeded	Wait and retry with exponential backoff
500	Internal Server Error	Server-side error	Retry after a brief delay
503	Service Unavailable	Model is temporarily unavailable	Check Model Status and retry

Error Response Format

{
  "detail": "Incorrect API key provided"
}

Note

Some endpoints (e.g., Azure-proxied models like GPT-4.1) may return OpenAI-style error objects:

{
  "error": {
    "message": "Invalid API key provided.",
    "type": "authentication_error",
    "code": "invalid_api_key"
  }
}

Handling Rate Limits (429)

When you hit a rate limit, implement exponential backoff:

import time
from openai import OpenAI, RateLimitError

client = OpenAI()

def call_with_retry(messages, max_retries=3):
    for attempt in range(max_retries):
        try:
            return client.chat.completions.create(
                model="Llama-3.3-70B-Instruct",
                messages=messages,
            )
        except RateLimitError:
            wait = 2 ** attempt
            print(f"Rate limited. Retrying in {wait}s...")
            time.sleep(wait)
    raise Exception("Max retries exceeded")

Common Issues

”Model not found”

• Verify the model name matches exactly (case-sensitive)
• Use the List Models API to check available models
• Some models may only be available on certain plan tiers

”Invalid API key”

• Ensure the key is correctly set in your environment
• Check that the key hasn’t expired
• Visit the API Key Portal to manage keys

”Context length exceeded”

• Reduce the input length or set a lower max_tokens
• Check the model’s max_sequence_length in the model metadata