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.

Using Roo Code with NanoGPT

A quick guide to setting up Roo Code with NanoGPT’s API.

Setup Instructions

  1. Get your API key from nano-gpt.com/api
  2. In Roo Code settings, add a new Custom Model with these details:
Tip: Roo Code surfaces the thinking and response channels side-by-side. Pointing it at https://api.nano-gpt.com/api/v1thinking/ keeps that richer reasoning stream intact for models like Minimax M2 or Kimi K2 Thinking.

Subscribers

With the Base URL above, Roo Code lists every NanoGPT model, including models your subscription doesn’t cover. To see only the models your subscription includes, use https://api.nano-gpt.com/api/subscription/v1/ as the Base URL instead: Roo Code then loads its model list from /api/subscription/v1/models. The subscription Base URL has no v1thinking variant, so it uses the standard reasoning stream shape. To keep a key from ever billing your balance, set its billing mode to Subscription only (see API key billing settings); requests your subscription doesn’t cover are then rejected instead of charged. That’s it! You can now use Roo Code with every model you can think of. Model names are on our pricing page - important ones are openai/gpt-6.1-sol, anthropic/claude-opus-5.5, google/gemini-flash-latest, or just any other model you fancy.