nano-gpt-plugin

NanoGPT provider plugin for OpenClaw

Features

• Dynamic model catalog from https://nano-gpt.com/api/v1/models?detailed=true
• Auto-populates model capabilities (vision, reasoning, context window, pricing)
• Handles NANOGPT_API_KEY environment variable
• Provides openclaw onboard --nano-gpt-api-key <key> flow
• Includes include_usage: true on all outgoing requests for accurate token tracking
• Subscription usage tracking via /api/subscription/v1/usage
• Balance checking via /api/check-balance
• Supports all NanoGPT model families: OpenAI, Anthropic, Google, xAI, DeepSeek, Moonshot, Qwen, Groq, and 50+ more

Installation

From clawhub

openclaw plugins install clawhub:@forceconstant/nano-gpt
openclaw onboard --non-interactive --nano-gpt-api-key "$NANOGPT_API_KEY" --flow quickstart

• Note for multi-agent setups, you will need to copy the auth-profile.json to each agent.
• Recommend also deleting each agents models.json to make sure all models are up to date.

Local development

Clone this repository and link it:

git clone <this-repo>
cd nano-gpt-plugin
pnpm install
pnpm run build

Then, in your OpenClaw workspace, you can use the plugin by referencing the built dist directory.

Usage

Onboarding

To add your NanoGPT API key:

openclaw onboard --nano-gpt-api-key <your-key>

This will store the key in your OpenClaw configuration.

Using models

Once onboarded, you can use any model from NanoGPT by referencing it with the nano-gpt/ prefix:

openclaw chat --model nano-gpt/openai/gpt-5.2 "Hello, world!"

Or in your agent configuration:

{
  "model": "nano-gpt/anthropic/claude-opus-4.6"
}

Usage and balance

To check your usage and balance:

openclaw status

This will show your daily/monthly token usage and remaining balance.

Configuration

The plugin does not require any additional configuration beyond the API key.

Development

Building

pnpm run build

Testing

pnpm test

Linting

pnpm run lint

License

MIT