Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Scan for outdated or missing drivers - takes under a minute3Repair Windows errors before they cause bigger problemsComfy MCP is a local server that lets Claude Code or Cursor inspect a ComfyUI installation on your own machine and build, run, and manage workflows in it. The setup has four parts: install comfy-mcp and comfy-cli version 1.14.0 or newer, make sure a ComfyUI workspace exists, start ComfyUI with comfy launch, and register comfy-mcp in your client. Then restart or reload the client so it picks up the new server.
Before you start
- Python 3.10 or newer, as stated in Comfy’s setup guidance for the MCP server.
- comfy-cli 1.14.0 or newer, available to the process that launches the MCP server. The server calls the
comfybinary found on yourPATH, or the path you set inCOMFY_BIN. It does not bundle or install comfy-cli for you. - A ComfyUI installation or workspace. If you do not have one,
comfy installcreates it. - ComfyUI running before you ask the agent to execute anything. The project’s guidance is to start it with
comfy launchand leave it running.
The 1.14.0 floor comes from the project’s documentation. Check the repository’s current README before installing, because the minimum version can change. The server checks the version of the binary it actually invokes, so an older comfy-cli elsewhere on your system does not count.
As an Amazon Associate I earn from qualifying purchases.
Install the packages and start ComfyUI
Installing comfy-mcp does not pull in comfy-cli, so install both. The commands below use pip; if you manage Python tools with uv or another installer, use the equivalent.
-
Install the MCP server and the CLI in the same Python environment:
#1 Best Overall
ASRock Intel Arc Pro B60 Creator 24GB Graphics Card, Workstation GPU, Xe2-HPG, 2400MHz, 24GB GDDR6 192-bit, PCIe 5.0, 4X DP 2.1, Blower- System Compatibility Note: 2-slot card, 271x112x39mm, single 8-pin power, 200W TDP. Verify chassis clearance and PSU capacity before purchase.
- Dedicated Support: Please contact us directly through Amazon for any product questions or assistance you may require.
- 24GB GDDR6 on 192-Bit Bus: Massive 24GB memory with 456 GB/s bandwidth – ideal for LLMs, AI inference, 3D rendering, and generative design.
- Intel Xe2-HPG Architecture: Built on Intel's next-gen architecture with 20 Xe cores and 160 XMX engines for AI acceleration (197 INT8 TOPS).
- PCIe 5.0 Support: PCI Express 5.0 x16 interface for maximum bandwidth with the latest workstation platforms.
pip install comfy-mcp "comfy-cli>=1.14.0" -
If you do not already have a ComfyUI workspace, create one:
comfy install -
Start ComfyUI and leave the terminal session running:
comfy launch -
In the same terminal, confirm that
comfyresolves. If the shell cannot find it, fix the environment before configuring a client, because the client will inherit the same problem.Free tools Windows power users keep installed
One-click scans. No signup required.
Special offer. See more information about Outbyte and uninstall instructions. Please review EULA and Privacy policy. -
Register
comfy-mcpwith your client using one of the two procedures below, then restart or reload the client.Rank #2
Sapphire 32358-01-20G AMD Radeon™ AI PRO R9700 Graphics Card with 32GB GDDR6, AMD RDNA 4- Memory Size: 32 GB, 256-bit GDDR6
- Output: 4 x DisplayPort 2.1a
- Interface: PCI-Express 5.0 x16
- Boost Clock: Up to 2920MHz
- Game Clock: Up to 2350 MHz
Connect Claude Code
Claude Code can register the server from the command line or read a project-level configuration file.
Register from the command line
The project’s example registers a local stdio server whose command is comfy-mcp:
claude mcp add comfy-mcp -- comfy-mcp
To pass environment variables to the server, add them to the registration using the CLI’s environment option. The project documents COMFY_BIN for a non-default comfy location and COMFY_API_KEY for partner-API nodes. The API key is optional and only matters if you use those nodes.
Quick wins for a faster PC:
Clear out junk files and repair common Windows errorsFree Scan →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Commit a project configuration file
To share the setup with a repository, add a .mcp.json file at the project root with the same server command:
Rank #3
- NVIDIA GeForce RTX 5080; Blackwell Architecture
- 16 GB GDDR7 RAM (memory interface)
- Core/memory clock: (boost: 2805) MHz
- 1x HDMI, 3x DP, height: dual slot
- DirectX 12
{
"mcpServers": {
"comfy-mcp": {
"command": "comfy-mcp"
}
}
}
Teammates who need a specific comfy path can add an env block with COMFY_BIN set to their own absolute path, rather than committing a path that exists only on your machine.
Connect Cursor
Cursor reads MCP servers from a JSON file with an mcpServers object. You can use a global file or a project file:
- Global:
~/.cursor/mcp.json, which applies to every project you open. - Project:
.cursor/mcp.jsoninside the project folder, which applies only there.
The entry is the same as the Claude Code one:
{
"mcpServers": {
"comfy-mcp": {
"command": "comfy-mcp",
"env": {
"COMFY_BIN": "/absolute/path/to/comfy"
}
}
}
}
Keep the env block only if Cursor cannot see the correct comfy executable. Replace the placeholder path with the output of which comfy from a terminal on macOS or Linux, or the equivalent lookup on Windows. Then reload Cursor so the MCP settings take effect.
Point the server at the right comfy binary
The server needs the comfy program to start ComfyUI and manage it. Desktop apps often launch with a smaller PATH than your terminal, so a command that works in your shell may fail inside Claude Code or Cursor. Setting COMFY_BIN to an absolute path removes that dependency. Use the path to the binary, not the directory that contains it.
Rank #4
- Chipset: NVIDIA GeForce GT 1030
- Video Memory: 4GB DDR4
- Boost Clock: 1430 MHz
- Memory Interface: 64-bit
- Output: DisplayPort x 1 (v1.4a) / HDMI 2.0b x 1
What the agent can do with it
The project lists these capabilities:
- Running workflows and generating images from a text prompt.
- Monitoring asynchronous jobs, cancelling them, and retrieving outputs.
- Listing the nodes, models, and templates available in your installation, including custom nodes.
- Validating workflows and editing templates.
- Managing the ComfyUI process and reading its logs.
The project labels the server beta and reported 40 tools in its documentation at the time of writing. Those numbers describe the server, not the quality of any result. Actual outcomes depend on your ComfyUI installation, the workflow’s dependencies, and the agent you connect.
Comfy’s announcement describes using hardware inspection to judge whether a model is suitable before you download it, and checking installed nodes and models when preparing a workflow. The announcement’s example model does not establish a minimum hardware requirement or a recommended GPU, so use the agent’s suitability check for your own machine rather than a generic spec.
Local ComfyUI or Comfy Cloud
Comfy’s support material describes two separate MCP connections. Do not confuse them: the local server is a subprocess that drives your ComfyUI, while the Cloud connection is a remote HTTP service that runs workflows on Comfy Cloud GPUs. They authenticate separately, and you can use both.
Crashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minutePC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11| Factor | Local Comfy MCP (comfy-mcp) | Comfy Cloud MCP |
|---|---|---|
| Where workflows execute | Your own machine | Comfy Cloud GPUs |
| Requires a local ComfyUI install | Yes; comfy install can create one |
Not stated in the support material as a requirement |
| Access to local models, files, and custom nodes | Yes | Not stated |
| Local hardware dependence | Yes; generation uses your GPU and resources | No local GPU use |
| Generation charges | Local generation does not incur Cloud-generation charges, per Comfy’s support article; partner models use credits | Requires an active subscription, per Comfy’s support article |
| Best fit | Existing local setups, private or custom nodes, local models | Avoiding a local install, or new users and browser-based assistants |
Commercial terms change, so confirm current Comfy pricing and subscription rules before you rely on the cost column.
Best Value
- Robust 4GB Memory & Quad Display Ready: Equipped with 4GB of fast GDDR5 memory to smoothly handle daily graphics tasks. Features four built-in HDMI ports, enabling a seamless quad-monitor setup directly out of the box—perfect for multi-tasking offices, digital signage, or trading desks.
- Plug-and-Play Installation & Wide Compatibility: Utilizes a standard PCI Express interface for broad compatibility with most desktop PCs. Offers straightforward plug-and-play installation and stable driver support for modern Windows and Linux operating systems, ensuring a hassle-free setup.
- Quiet, Cool & Compact Design: Engineered with a silent fan and efficient cooling system for near-silent operation, making it ideal for noise-sensitive environments. Its low-profile design fits easily into small form factor cases, with both half-height and full-height brackets included for flexible installation.
- Enhanced Multimedia & Everyday Performance: Delivers smooth 1080P video playback and supports hardware-accelerated decoding, offering an excellent experience for home theater PCs (HTPC). Provides capable performance for everyday applications, multimedia tasks.
- Complete Package & Reliable Support: Includes the graphics card, both low-profile and standard brackets, a quick start guide, and screwdriver, which make it simple and quick setup process.
Troubleshooting
-
The server connects, but tool calls report that
comfyis missing. Install comfy-cli in the same environment as the server, then setCOMFY_BINto the absolute path of the binary. -
No Comfy tools appear after configuration. Confirm the JSON is valid and the entry is saved, then restart or reload the client.
-
On macOS, the server fails when the installation sits in Documents, Desktop, or Downloads. Privacy protections can block access. Move the installation to a directory outside those folders. Alternatively, open System Settings, go to Privacy & Security, then Full Disk Access, and grant it to the app that launches the server, such as your terminal or Cursor. Restart that app afterwards.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Special offer. See more information about Outbyte and uninstall instructions. Please review EULA and Privacy policy. -
Workflow calls fail immediately. Check that ComfyUI is running. Start it with
comfy launchif it is not. -
The connection works, but a workflow will not run. The server being connected does not mean every workflow has what it needs. Required nodes, models, and dependencies must be present in the ComfyUI environment the server targets. Ask the agent to list the installed nodes and models, and to validate the workflow, to find the gap.
Prompts that work as a starting point
Comfy’s announcement gives a setup request you can use directly: “Help me set up the local Comfy MCP connection.” For a first run against an existing workflow, the project’s example is:
- “Confirm my local ComfyUI is running, then run the workflow at
~/workflows/txt2img.jsonand show me the image.”
The first confirms that the local service is reachable. The second tests the full path from the agent to ComfyUI and back.
The Tool Desk
Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.




