Connect Models

LingXiao supports multiple LLM providers, configured via environment variables or config files.

Supported Providers

OpenAI (and compatible: DeepSeek, Qwen, Moonshot/Kimi, Gemini, Groq, SiliconFlow, etc.)
Anthropic

Environment Variables

Variable	Description
`LINGXIAO_LLM_PROVIDER`	`auto` / `openai` / `anthropic`
`LINGXIAO_OPENAI_API_KEY`	OpenAI or compatible API key
`LINGXIAO_OPENAI_BASE_URL`	OpenAI compatible endpoint
`LINGXIAO_ANTHROPIC_API_KEY`	Anthropic key
`LINGXIAO_ANTHROPIC_BASE_URL`	Anthropic endpoint
`LINGXIAO_LEADER_MODEL`	Leader model
`LINGXIAO_AGENT_MODEL`	Worker model

Config File

Edit ~/.lingxiao/settings.json:

{
  "llm": {
    "provider": "openai",
    "openai": {
      "apiKey": "sk-xxx",
      "baseUrl": "https://api.openai.com/v1"
    },
    "leader_model": "gpt-4o",
    "agent_model": "gpt-4o-mini"
  }
}

Using Compatible Endpoints

DeepSeek example:

export LINGXIAO_LLM_PROVIDER=openai
export LINGXIAO_OPENAI_API_KEY=sk-xxx
export LINGXIAO_OPENAI_BASE_URL=https://api.deepseek.com/v1
export LINGXIAO_LEADER_MODEL=deepseek-chat
export LINGXIAO_AGENT_MODEL=deepseek-chat

Anthropic example:

export LINGXIAO_LLM_PROVIDER=anthropic
export LINGXIAO_ANTHROPIC_API_KEY=sk-ant-xxx
export LINGXIAO_LEADER_MODEL=claude-sonnet-4-20250514
export LINGXIAO_AGENT_MODEL=claude-sonnet-4-20250514

Local LLM Gateway

LingXiao includes a local LLM gateway that serves as an OpenAI/Anthropic dual-format proxy:

Virtual key management
Per-key RPM/TPM/daily token budget rate limiting
Request trace logging to llm_gateway_requests table
Default port: 62000

Enable in ~/.lingxiao/settings.json:

{
  "llm_gateway": {
    "enabled": true,
    "provider": "openai",
    "model": "gpt-4o",
    "default_rpm": 60,
    "default_tpm": 200000,
    "trace_enabled": true
  }
}

Next Steps

First Run — Launch your first expert panel