openclaw-config

Standalone CLI to add or update OpenClaw config (e.g. models.providers) without depending on the openclaw package. Only edits openclaw.json; no coupling to openclaw source.

Install

npm install -g openclaw-config
# or
npx openclaw-config add-provider ...

Usage

Add a model provider

openclaw-config add-provider <providerId> --base-url <url> --models <id1,id2,...> [options]

Options:

| Option | Description | | ------------ | ------------------------------------------------------------------------ | | --base-url | Provider base URL (required) | | --api-key | API key (optional) | | --api | openai-completions | openai-responses (default: openai-completions) | | --config | Path to openclaw.json (default: env or ~/.openclaw/openclaw.json) |

Config path resolution (same as OpenClaw):

1. OPENCLAW_CONFIG_PATH → use that file
2. OPENCLAW_STATE_DIR → <dir>/openclaw.json
3. Default: ~/.openclaw/openclaw.json

Examples

NVIDIA NIM:

openclaw-config add-provider nvidia \
  --base-url "https://integrate.api.nvidia.com/v1" \
  --api-key "your-api-key" \
  --models "minimaxai/minimax-m2.1,z-ai/glm4.7"

LM Studio (local, default base URL):

openclaw-config add-provider lmstudio \
  --base-url "http://localhost:1234/v1" \
  --models "qwen/qwen3-coder-next"

Ollama:

openclaw-config add-provider ollama \
  --base-url "http://127.0.0.1:11434/v1" \
  --models "llama3.2"

xAI (Grok):

openclaw-config add-provider xai \
  --base-url "https://api.x.ai/v1" \
  --api-key "your-api-key" \
  --models "grok-beta"

Minimax:

openclaw-config add-provider minimax \
  --base-url "https://api.minimax.chat/v1" \
  --api-key "your-api-key" \
  --models "MiniMax-M2.1"

Moonshot (Kimi):

openclaw-config add-provider moonshot \
  --base-url "https://api.moonshot.ai/v1" \
  --api-key "your-api-key" \
  --models "kimi-k2.5"

Qwen Portal:

openclaw-config add-provider qwen \
  --base-url "https://portal.qwen.ai/v1" \
  --api-key "your-api-key" \
  --models "qwen-plus"

Qianfan (Baidu):

openclaw-config add-provider qianfan \
  --base-url "https://qianfan.baidubce.com/v2" \
  --api-key "your-api-key" \
  --models "deepseek-v3.2"

Xiaomi:

openclaw-config add-provider xiaomi \
  --base-url "https://api.xiaomimimo.com/anthropic" \
  --api-key "your-api-key" \
  --models "mimo-v2-flash"

Synthetic:

openclaw-config add-provider synthetic \
  --base-url "https://api.synthetic.new/anthropic" \
  --api-key "your-api-key" \
  --models "synthetic-model"

OpenRouter:

openclaw-config add-provider openrouter \
  --base-url "https://openrouter.ai/api/v1" \
  --api-key "your-api-key" \
  --models "openrouter/pony-alpha"

Provider that only supports /v1/responses:

openclaw-config add-provider codex \
  --base-url "https://api.134300.xyz/v1" \
  --api openai-responses \
  --api-key "your-api-key" \
  --models "gpt-5.3-codex,gpt-5.2"

Programmatic API

import { addProvider } from "openclaw-config";

const configPath = addProvider({
  providerId: "nvidia",
  baseUrl: "https://integrate.api.nvidia.com/v1",
  apiKey: "your-api-key",
  models: ["minimaxai/minimax-m2.1", "z-ai/glm4.7"],
});
console.log("Written to", configPath);

License

MIT