OpenAI¶

Configure HolmesGPT to use OpenAI's GPT models.

Setup¶

Note

This requires a paid OpenAI API key, not a ChatGPT Plus subscription.

Configuration¶

Using CLI Arguments¶

holmes ask --api-key="your-openai-api-key" "what pods are crashing in my cluster and why?"

Using Environment Variables¶

export OPENAI_API_KEY="your-openai-api-key"
holmes ask "what pods are crashing in my cluster and why?"

Available Models¶

OpenAI is the default provider. Specify models with the --model flag:

# Default model (gpt-4o)
holmes ask "what pods are failing?"

# Specific models
holmes ask "what pods are failing?" --model="gpt-4"
holmes ask "what pods are failing?" --model="gpt-3.5-turbo"

Troubleshooting¶

Invalid API Key

Error: Invalid OpenAI API key

- Verify your API key is correct - Ensure your account has sufficient credits

Rate Limits

Error: Rate limit exceeded

- Wait for the limit to reset or upgrade your OpenAI account

Insufficient Credits

Error: You exceeded your current quota

- Add payment method to your OpenAI account