# Onde Inference CLI: Fine-tune & Deploy Local LLMs

> Onde Inference CLI: Fine-tune & Deploy Local LLMs is an open-source MCP server for ai models and llm tooling. Fine-tune Qwen models with LoRA, export GGUF and manage Onde apps from an agent. It runs locally on your machine and is started by your AI client.

- **Kind:** MCP server
- **Pricing:** Open source
- **Platforms:** CLI, macOS, Linux, Windows
- **Website:** https://ondeinference.com/cli
- **Repository:** https://github.com/ondeinference/onde-cli
- **MCP Registry name:** io.github.ondeinference/onde-cli
- **Categories:** AI models and LLM tooling, developer tools
- **Listing updated:** 2026-11-14

## About

onde is the command-line client for Onde Inference, and running it with the MCP flag serves its functions to an agent over stdio. The same binary opens a terminal UI for signing in and managing apps and the model catalog.

Its local workflow picks a safetensors model, starts a LoRA run on your machine, merges the adapter into the base model and exports GGUF for local inference or packaging. The README names Qwen models as the supported fine-tuning targets. Training data, model files and logs stay in one place on your computer.

It installs through Homebrew, npm, Cargo, pip or uv, NuGet, Dart pub and prebuilt binaries. The code is dual-licensed under MIT and Apache 2.0. The wider Onde platform has separate paid options, such as per-device plans and a pay-per-token cloud endpoint, which are not needed to use the CLI.

## Features

- **Terminal UI:** Sign in and manage your account, apps and models from the terminal.
- **Local LoRA tuning:** Fine-tune supported Qwen models on your own machine.
- **Adapter merge:** Merge the adapter back into the base model.
- **GGUF export:** Export the merged model for local inference.
- **MCP mode:** Run as an MCP server over stdio for agents.
- **Many install routes:** Homebrew, npm, Cargo, pip, NuGet, Dart pub and binaries.

## Good for

- Fine-tune a small model on a Mac
- Export a model to GGUF for an app
- Let an agent manage Onde apps and models
- Prepare a model to ship on-device

## Install

### Claude Code (Terminal)

```sh
claude mcp add onde-cli -- npx -y @ondeinference/cli
```

### Claude Desktop (claude_desktop_config.json)

```json
{
  "mcpServers": {
    "onde-cli": {
      "command": "npx",
      "args": [
        "-y",
        "@ondeinference/cli"
      ]
    }
  }
}
```

### Cursor (~/.cursor/mcp.json)

```json
{
  "mcpServers": {
    "onde-cli": {
      "command": "npx",
      "args": [
        "-y",
        "@ondeinference/cli"
      ]
    }
  }
}
```

### VS Code (.vscode/mcp.json)

```json
{
  "servers": {
    "onde-cli": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "@ondeinference/cli"
      ]
    }
  }
}
```

### Windsurf (~/.codeium/windsurf/mcp_config.json)

```json
{
  "mcpServers": {
    "onde-cli": {
      "command": "npx",
      "args": [
        "-y",
        "@ondeinference/cli"
      ]
    }
  }
}
```

### Other clients (your client's MCP settings (most use this shape))

```json
{
  "mcpServers": {
    "onde-cli": {
      "command": "npx",
      "args": [
        "-y",
        "@ondeinference/cli"
      ]
    }
  }
}
```

## FAQ

### How do I start the MCP server?

Run onde --mcp for a stdio server.

### Which models can I tune?

The README says Qwen models, using LoRA.

### What license is it?

Dual-licensed under MIT and Apache 2.0.

### Does the CLI cost money?

Install pricing is not listed; Onde's paid plans apply to its device fleet and cloud endpoint.

### Is Onde Inference CLI: Fine-tune & Deploy Local LLMs free?

Yes. Onde Inference CLI: Fine-tune & Deploy Local LLMs is open source, so you can use and self-host it for free.

### How do I add Onde Inference CLI: Fine-tune & Deploy Local LLMs to Claude Code?

Run this command in your terminal: claude mcp add onde-cli -- npx -y @ondeinference/cli

### What are the best alternatives to Onde Inference CLI: Fine-tune & Deploy Local LLMs?

The closest alternatives to Onde Inference CLI: Fine-tune & Deploy Local LLMs are REA, Codebase Memory, GitHub, Serena MCP: the IDE for your agent and XERJ.

## Alternatives

- [REA](https://ai-powered.com/mcp/rea): REA is an open-source MCP server for developer tools.
- [Codebase Memory](https://ai-powered.com/mcp/codebase-memory): Codebase Memory is an open-source MCP server for knowledge and memory.
- [GitHub](https://ai-powered.com/mcp/github-2): GitHub is an open-source MCP server for developer tools.
- [Serena MCP: the IDE for your agent](https://ai-powered.com/mcp/serena-mcp-the-ide-for-your-agent): Serena MCP: the IDE for your agent is an open-source MCP server for developer tools.
- [XERJ](https://ai-powered.com/mcp/xerj): XERJ is an open-source MCP server for knowledge and memory.
- [Ouroboros](https://ai-powered.com/mcp/ouroboros): Ouroboros is an open-source MCP server for developer tools.
- [Claude Code Ultimate Guide](https://ai-powered.com/mcp/claude-code-ultimate-guide): Claude Code Ultimate Guide is an open-source MCP server for ai models and llm tooling.
- [Concord MCP](https://ai-powered.com/mcp/concord-mcp): Concord MCP is a freemium MCP server for developer tools.

---
Source: https://ai-powered.com/mcp/onde-inference-cli-fine-tune-and-deploy-local-llms
