Skip to content
relaxAI Docs

Troubleshooting

Common Issues

1. Authentication Error:

  • Verify your API key is correct
  • Ensure the API key is properly formatted in the authorization header
  • Check for any whitespace in the API key
  • Confirm your account has proper permissions

2. Model Not Found:

  • Double-check the model name spelling (case-sensitive)
  • Verify the model is available in relaxAI
  • Ensure you’re using the fully qualified model name
  • Try using another available model as a test

3. Context Window Exceeded:

  • Reduce input tokens by summarizing or chunking
  • Split requests into smaller, manageable chunks
  • Implement proper token counting to stay within limits
  • Remember: The context window of DeepSeek-R1-0528 is 65k tokens. For problems requiring large context windows, please use Llama-4-Maverick-17B-128E (1M token context window)

4. Slow Response Times:

  • DeepSeek-R1-0528 is a reasoning model and thus has longer response times. For fast responses, try Llama-4-Maverick-17B-128E.
  • Optimize your prompts to be more concise
  • Implement proper timeout handling in your application
  • Check network connectivity and latency

5. Integration-Specific Issues:

  • Ensure you’re using the correct API base URL format
  • Verify all required parameters are properly configured
  • Check integration-specific documentation for any special requirements
  • Look for error logs in the specific tool or framework

Debug Tips

For debugging API calls, use curl to test the API directly:

Terminal window
curl https://api.relax.ai/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer RELAX_API_KEY" \
  -d '{
    "model": "Llama-4-Maverick-17B-128E",
    "messages": [{"role": "user", "content": "Hello world"}]
  }'

Advanced Troubleshooting

1. Enable Verbose Logging:

  • Most integrations allow enabling debug or verbose logging:
  • LangChain: langchain.debug = True
  • Python requests: Use verbose=True or logging configurations
  • Node.js: Set DEBUG=openai* environment variable

2. Monitoring Tools:

  • Use API monitoring tools like Postman or Insomnia for request inspection
  • Implement logging middleware to capture request/response details
  • Add timing metrics to identify performance bottlenecks

3. Rate Limiting Issues:

  • Implement proper backoff and retry logic
  • Monitor rate limit headers in API responses
  • Add request queuing for high-volume applications