Use SimpleLLM — EU-hosted, DSGVO-compliant LLM inference — as a model provider in OpenClaw. One command sets everything up.
SimpleLLM exposes an OpenAI-compatible API. This package generates the OpenClaw provider config (model list, pricing, context windows) and stores your API key, so openclaw works out of the box.
npx @simplellm/openclaw-providerThe setup wizard connects to SimpleLLM, fetches the available models, asks for your API key, and writes:
~/.openclaw/openclaw.json— provider + model config~/.openclaw/.env— yourSIMPLELLM_API_KEY
Then just run:
openclawThe default model is simplellm/Qwen3-Coder-30B-A3B-Instruct.
Re-run any time models change — it merges into your existing config rather than replacing it.
Pass the key as an argument or via env for scripted setups:
npx @simplellm/openclaw-provider sk-...
# or
SIMPLELLM_API_KEY=sk-... npx @simplellm/openclaw-provider| Variable | Default | Purpose |
|---|---|---|
SIMPLELLM_BASE_URL |
https://api.simplellm.eu |
Point at a self-hosted or local SimpleLLM instance |
OPENCLAW_STATE_DIR |
~/.openclaw |
Where the config and .env are written |
SIMPLELLM_API_KEY |
— | Provide the API key non-interactively |
Get an API key at simplellm.eu.
Requires Node.js 20+.
git clone https://github.com/SimpleLLM/openclaw-provider.git
cd openclaw-provider
npm install
npm run build # tsc → dist/- @simplellm/opencode-provider — the same, for OpenCode
MIT — see LICENSE.