Power your OpenClaw agent with open-source models

    Configure Kimi K2.5, GLM 5, and MiniMax M2.5 as your OpenClaw inference provider — in 30 seconds. No Anthropic or OpenAI API key needed.

    terminal

    How It Works

    Step 1

    Install Kimchi

    bash
    curl -fsSL https://github.com/getkimchi/kimchi/releases/latest/download/install.sh | bash

    One command. Works on macOS and Linux.

    Step 2

    Run the wizard

    bash
    kimchi

    Kimchi auto-detects OpenClaw and configures it.

    Step 3

    Build with your agent

    Your OpenClaw agent now runs on open-source models via Cast AI.

    Three models. Zero vendor lock-in.

    Curated for agentic workloads — large context windows, tool-use optimized, open-source.

    Kimi K2.5

    Recommended

    Primary Agent

    Reasoning, planning, code generation, image processing

    Context: 262K
    Output: 32K

    GLM 5 FP8

    Coding Subagent

    Writing, refactoring, debugging code

    Context: 202.8K
    Output: 32K

    MiniMax M2.5

    Budget-friendly

    Secondary Subagent

    Code generation, debugging

    Context: 196.6K
    Output: 32K

    Why not just use OpenRouter or Anthropic directly?

    FeatureKimchiOpenRouterAnthropic DirectOllama
    Setup time30 seconds5–10 min5 min30+ min
    Open-source modelsCuratedHundredsProprietarySelf-host
    One API keyLocal
    Agent-optimizedLarge ctx + toolsGenericExpensiveHW-limited
    Free tierFree (your GPU)
    Auto-configBuilt-inManual JSONManualManual

    What Kimchi writes to your openclaw.json

    bash
    // ~/.openclaw/openclaw.json
    {
      "models": {
        "providers": {
          "kimchi": {
            "baseUrl": "https://llm.kimchi.dev/openai/v1",
            "apiKey": "${KIMCHI_API_KEY}",
            "api": "openai-completions",
            "models": [
              {
                "id": "kimchi/kimi-k2.5",
                "name": "Kimi K2.5",
                "reasoning": true,
                "contextWindow": 262000,
                "maxTokens": 32000
              },
              {
                "id": "kimchi/glm-5-fp8",
                "name": "GLM 5 FP8",
                "reasoning": true,
                "contextWindow": 202800,
                "maxTokens": 32000
              },
              {
                "id": "kimchi/minimax-m2.5",
                "name": "MiniMax M2.5",
                "reasoning": false,
                "contextWindow": 196600,
                "maxTokens": 32000
              }
            ]
          }
        }
      }
    }

    Kimchi preserves your existing OpenClaw config. It only adds the kimchi provider — your other providers, channels, and agent settings stay untouched.

    Start in 30 seconds

    bash
    curl -fsSL https://github.com/getkimchi/kimchi/releases/latest/download/install.sh | bash
    1. 1Run the install command
    2. 2Get your free API key at app.kimchi.dev
    3. 3Run kimchi and select OpenClaw
    4. 4Done — your agent is ready
    View on GitHub

    Start building with open-source models.

    Get started free — no credit card required.

    Start for free