Skip to content
MCP server

Unity MCP

AI agents inspect and edit a running Unity Editor: scenes, prefabs, tests and builds

Runs locallyStatus and history
Pricing
Open source
Platforms
Windows, macOS, Linux
Runs
Locally
Version
4.4.2
Listing updated
1 day ago

About Unity MCP

Unity MCP is an open-source MCP server for developer tools. It runs locally on your machine and is started by your AI client.

There is no separate server process. The Unity Editor itself exposes a local Streamable HTTP endpoint and writes a descriptor file that the command line tool reads, so no port or token has to be configured. The README reports it working with Claude Code, Cursor, Codex, Gemini CLI, VS Code and Claude Desktop, and the project must stay open in the Editor while it is used.

Tools fall into diagnostics (console, Editor.log, compile state, tests, scene hierarchy, asset reads), authoring (GameObjects, components, assets, scenes, prefabs, Animator Controllers), rendering, Timeline and Recorder, builds, C# snippet execution and synthesised Editor input. Each authoring call is one Undo step. Timeline, Recorder and test tools appear only when the matching Unity packages are installed, and a URL parameter can limit the tool list to chosen groups. A verify command recompiles and collects errors in one call, optionally running tests.

New tools are added by putting an attribute on a C# static method. Installation is through the Package Manager from a Git URL or through VCC and ALCOM, and the code is MIT licensed. It needs Unity 2022.3 or newer; the documentation is in Japanese with English, and a no-command-line route uses a Claude Desktop extension.

Install

Run in a terminal

claude mcp add unity-mcp -- dnx IsuzuUnityCli --yes

Add to claude_desktop_config.json

{
  "mcpServers": {
    "unity-mcp": {
      "command": "dnx",
      "args": [
        "IsuzuUnityCli",
        "--yes"
      ]
    }
  }
}
Install in CursorOpens Cursor, which asks you to confirm. Or add it by hand:

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "unity-mcp": {
      "command": "dnx",
      "args": [
        "IsuzuUnityCli",
        "--yes"
      ]
    }
  }
}
Install in VS CodeOpens VS Code, which asks you to confirm. Or add it by hand:

Add to .vscode/mcp.json

{
  "servers": {
    "unity-mcp": {
      "type": "stdio",
      "command": "dnx",
      "args": [
        "IsuzuUnityCli",
        "--yes"
      ]
    }
  }
}

Add to ~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "unity-mcp": {
      "command": "dnx",
      "args": [
        "IsuzuUnityCli",
        "--yes"
      ]
    }
  }
}

Add to your client's MCP settings (most use this shape)

{
  "mcpServers": {
    "unity-mcp": {
      "command": "dnx",
      "args": [
        "IsuzuUnityCli",
        "--yes"
      ]
    }
  }
}

Configuration

  • UNITY_MCP_STATE_DIRWhere the Editor writes its descriptors. Set this only when the default location cannot be reached, such as from WSL.
  • UNITY_MCP_HOSTReplaces 127.0.0.1 in the descriptor endpoint, for reaching an Editor on the Windows host from WSL.

Tools this server provides (3)

Taken from the project documentation; the running server may differ.

play_mode_status
Reports whether the Editor is in play mode
test_run
Runs tests when the Unity test framework is present
test_results
Returns test results

What it does

In-Editor server
The Editor serves the MCP endpoint directly, with no separate process.
Diagnostics tools
Reads console, log, compile status, tests, hierarchy and assets.
Scene and asset authoring
Creates and changes GameObjects, components, prefabs, scenes and Animator Controllers.
Undo-friendly edits
Each authoring call collapses into one Undo operation.
Verify command
Recompiles, extracts errors and optionally runs tests in one call.
Custom tools
Add a tool by tagging a C# static method.
Tool groups
Limits the tool list to groups such as diagnostics or authoring.

Good for

  • Ask an agent to add objects to a scene and read back Console errors
  • Recompile after script edits and list the errors
  • Run EditMode tests from an agent
  • Automate Timeline or Recorder setup
  • Build a project from the command line

Pricing

Open source

Prices change. Check unity-mcp.shiranui-isuzu.dev before you buy.

Unity MCP FAQ

What Unity version is required?

Unity Editor 2022.3 or newer.

Do I need to run a separate server?

No. The Editor exposes the endpoint and starts it when the project opens.

Can I use it without a terminal?

Yes, one route uses a Claude Desktop extension after adding the package in Unity.

Which clients are verified?

Claude Code, Cursor, Codex, Gemini CLI, VS Code and Claude Desktop.

What licence is it?

MIT.

Is Unity MCP free?

Yes. Unity MCP is open source, so you can use and self-host it for free.

How do I add Unity MCP to Claude Code?

Run this command in your terminal: claude mcp add unity-mcp -- dnx IsuzuUnityCli --yes

What are the best alternatives to Unity MCP?

The closest alternatives to Unity MCP are REA, Codebase Memory, GitHub, Serena MCP: the IDE for your agent and XERJ.

Alternatives to Unity MCP

  1. REA

    MCP server

    CLI and MCP server that lets coding agents reverse engineer apps and binaries

    Runs locally
    • Open source
    • 464K GitHub stars
    • Developer tools
  2. Codebase Memory

    MCP server

    Local MCP server that builds a code knowledge graph for AI coding agents

    Runs locally
    • Open source
    • 48.7K GitHub stars
    • Knowledge and memory
  3. GitHub

    MCP server

    Official MCP server connecting AI assistants to GitHub repos, issues and workflows

    Not checked yet
    • Remote
    • Open source
    • 34.5K GitHub stars
    • Developer tools
  4. MCP toolkit giving coding agents symbol-level code retrieval and editing

    Runs locally
    • Open source
    • 31.5K GitHub stars
    • Developer tools
  5. XERJ

    MCP server

    Local search engine for AI agents with cited answers over your code and docs

    Runs locally
    • Open source
    • 19.2K GitHub stars
    • Knowledge and memory
  6. Ouroboros

    MCP server

    Spec-first layer that pins requirements and verifies results of AI coding agent work

    Runs locally
    • Open source
    • 6.4K GitHub stars
    • Developer tools