Back to GO AI Chat GO AI Chat · Model Context Protocol

Connect GO AI Chat
to any MCP client.

Turn on the on-device MCP server in GO AI Chat for Mac, and Claude Code, Cursor, or Claude Desktop can call its image, video, and media tools directly — every generation metered against your signed-in Lab-token balance.

Download on theMac App Store See the setup steps

The MCP server is built into GO AI Chat for Mac (Mac Catalyst). It runs on-device, binds to localhost only, and authenticates with a Bearer token — so you stay in control.

Setup

Connect in three steps

Under two minutes from a fresh install to your first generation from the command line.

  1. 1

    Enable the server

    Open GO AI Chat on your Mac, go to Settings → MCP Server, and switch on Enable MCP Server. The status dot turns green and shows the live endpoint http://127.0.0.1:8722/mcp. Change the port any time if 8722 is taken.

  2. 2

    Copy your pairing token

    In the Pairing Token section, tap Copy. This token authenticates every request — treat it like a password. Hit Regenerate to roll it and instantly cut off any client still using the old one.

  3. 3

    Add it to your client

    Paste the connection snippet below into your MCP client. Replace <YOUR_TOKEN> with the token you just copied (the in-app snippets already include it for you).

    Claude Code

    Run in a terminal:

    claude mcp add --transport http goai-chat http://127.0.0.1:8722/mcp --header "Authorization: Bearer <YOUR_TOKEN>"
    Cursor & JSON-config clients

    Add to mcp.json (Cursor) or claude_desktop_config.json (Claude Desktop):

    {
  "mcpServers": {
    "goai-chat": {
      "type": "http",
      "url": "http://127.0.0.1:8722/mcp",
      "headers": { "Authorization": "Bearer <YOUR_TOKEN>" }
    }
  }
}
Capabilities

Thirteen tools, one server

Image, video, and media generation powered by Grok, plus on-device Studio color grading and free read-only helpers. Call list_generation_options for the full machine-readable catalog.

Generate

generate_image

Text-to-image in a chosen style preset.

1 token
transform_image

Image-to-image restyle, including inpainting via a PNG mask.

1 token
generate_video

Text-to-video clip. Returns a job id to poll.

per second
summarize_media

Summarize a YouTube URL or a PDF/MP4 file (≤50 MB).

5 tokens

Studio

studio_create_image

Create an image from a prompt — no style preset.

1 token
studio_edit_image

Edit a reference image with a text prompt.

2 tokens
studio_image_to_video

Animate a reference image into a video clip.

per second
studio_edit_video

Edit an existing video (≤15 s) with a prompt.

per second
studio_adjust_image_colors

Professional color grading on an image, on-device.

1 token
studio_adjust_video_colors

Professional color grading on a video, on-device.

3 tokens

Utility free

get_token_balance

Your Lab/chat token balances, tier, and 720p eligibility.

Free
get_job_status

Check the status, progress, or result of an async job.

Free
list_generation_options

The full catalog: styles, aspect ratios, durations, costs.

Free
Pricing

What each call costs

Every generation spends from the same Lab-token balance you use inside the app. Read-only tools are always free.

ToolCost
generate_image, transform_image, studio_create_image1 token
studio_edit_image2 tokens
studio_adjust_image_colors1 token
studio_adjust_video_colors3 tokens
summarize_media5 tokens
Video — generate_video, studio_image_to_video, studio_edit_videomax(10, round(seconds × 2.5)) — e.g. an 8 s clip = 20 tokens
get_token_balance, get_job_status, list_generation_optionsFree

720p video requires an active subscription. Check your balance any time with get_token_balance.

Under the hood

Local, metered, and yours

Localhost only

The server binds to 127.0.0.1. Nothing is exposed to your network or the internet — clients connect from the same Mac.

Bearer-token auth

Every request carries your pairing token. Regenerate it to revoke access instantly — old clients stop the moment you roll it.

Metered by token

Generations draw from your signed-in Lab-token balance — the same one the app uses. Read-only tools never spend tokens.

Async video jobs

Video tools return a job id right away. Poll get_job_status until the clip is ready — no blocking calls.

FAQ

Common questions

Which clients can connect?

Any MCP client that supports the streamable HTTP transport with a Bearer Authorization header — Claude Code, Cursor, and Claude Desktop are tested. Other spec-compliant clients work the same way.

Do I need the GO AI Chat Mac app?

Yes. The server runs on-device inside GO AI Chat for Mac. Keep the app open while a client is connected; closing it stops the server.

Is it available on iPhone or iPad?

Not yet — the MCP server is Mac-only today. The same account and Lab-token balance back it, so anything you generate also lands in your history across devices.

What does it cost?

Connecting is free. Generation tools are metered against your Lab tokens (images 1, Studio edit-image 2, media summaries 5, video per second). The read-only tools — balance, job status, and the options catalog — are free.

How do I revoke a client?

Open Settings → MCP Server and tap Regenerate on the pairing token, or switch the server off. Either one cuts existing clients off immediately.

The default port is taken — can I change it?

Yes. Set any port ≥ 1024 in the MCP Server settings. If the chosen port is busy the server automatically tries the next few until one is free, then shows the live URL.

Ready to wire it up?

Get GO AI Chat for Mac, flip on the MCP server, and call it from your favorite client.

Download on theMac App Store Explore GO AI Chat

Questions? support@goaichats.com