Skip to content

hedging8563/tokenlab-openclaw-provider

Repository files navigation

TokenLab Provider for OpenClaw

Use TokenLab models in OpenClaw through the normal OpenClaw inference loop.

The plugin uses TokenLab's OpenAI-compatible Chat Completions transport for agent execution. It refreshes chat model metadata from TokenLab's public /v1/models catalog, including pricing, context windows, output limits, vision support, reasoning, and tool-use capability. A curated 18-model catalog remains available when live discovery is unavailable.

Install

openclaw plugins install @tokenlabai/openclaw-provider
openclaw onboard --tokenlab-api-key "$TOKENLAB_API_KEY"

Or install the npm package first:

npm install -g @tokenlabai/openclaw-provider

The default model is tokenlab/gpt-5.5. Any model ID accepted by TokenLab can also be addressed as tokenlab/<model-id> through OpenClaw's dynamic model resolution.

Configuration

Set the API key before starting OpenClaw:

export TOKENLAB_API_KEY="your-tokenlab-api-key"

The default base URL is https://api.tokenlab.sh/v1. OpenClaw uses the OpenAI-compatible chat route. TokenLab also exposes native OpenAI Responses, Anthropic Messages, Gemini generateContent, image, video, music, 3D, audio, embeddings, rerank, translation, files, and async task endpoints for clients that support those contracts.

Development

npm install
npm run check
npx clawhub package validate .

Links

About

TokenLab model provider plugin for OpenClaw with live catalog discovery

Topics

Resources

License

Stars

0 stars

Watchers

0 watching

Forks

Packages

 
 
 

Contributors