Skip to main content
These examples use the direct API host, https://api.nano-gpt.com. It has no IPv6 address, so clients with only IPv6 should use https://nano-gpt.com with the same paths. See API Hosts.

LiteLLM Integration

Configure LiteLLM to proxy requests through NanoGPT while preserving the legacy reasoning_content field expected by its OpenAI-compatible connector.

Quick configuration

Add NanoGPT to your litellm.yaml (or the equivalent configuration source):
Set your API key before launching LiteLLM:
LiteLLM’s OpenAI adapter expects streaming deltas in delta.reasoning_content, so the v1legacy endpoint is the recommended base URL. If you upgrade your LiteLLM deployment to parse the modern delta.reasoning field, you can switch the base URL to https://api.nano-gpt.com/api/v1/ instead. You can also swap model to anthropic/claude-opus-5.5 or google/gemini-flash-latest if you want a different default.