Server#

gptme has a minimal REST API with very minimalistic web UI.

Note

The server and web UI is still in development and does not have all the features of the CLI. In particular, it doesn’t ever ask for confirmation before executing (yet).

It can be started by running the following command:

gptme-server

For more CLI usage, see the CLI reference.

There are a few different interfaces available:

Web UI#

A tiny chat interface with minimal dependencies that is bundled with the server.

Simply start the server to access the interface at http://localhost:5000

Fancy Web UI#

A modern, feature-rich web interface for gptme is available as a separate project gptme-webui. It is being built with gptengineer.app.

You can access it directly at gptme.gptengineer.run.

To serve gptme-webui locally, see the README.

Features:

  • Modern React-based interface with shadcn/ui

  • Streaming of responses

  • Mobile-friendly design

  • Dark mode support

  • Offline/exports support

  • Computer use interface integration

Computer Use Interface#

The computer use interface provides a split view with a chat on the left and a desktop on the right.

Requires Docker.

Warning

The computer use interface is experimental and has serious security implications. Please use with caution and see Anthropic’s documentation on computer use for additional guidance.

To run the computer use interface in Docker, follow these steps:

# Clone the repository
git clone https://github.com/ErikBjare/gptme.git
cd gptme
# Build container
make build-docker-computer
# Run container
docker run -v ~/.config/gptme:/home/computeruse/.config/gptme -p 6080:6080 -p 8080:8080 gptme-computer:latest

The computer use interface provides:

Features:

  • Split view with chat on the left, desktop on the right

  • Toggle for view-only/interactive desktop mode

  • Automatic screen scaling for optimal LLM vision

Requirements:

  • Docker for running the server with X11 support

  • Network ports 6080 (VNC) and 8080 (web UI) available

Using Computer Use Locally

You can use the computer tool (which enables computer use) locally on Linux without Docker or VNC, but it is not recommended due to security concerns.

Requirements: - X11 server - xdotool installed

To enable the computer tool, specify a -t/--tools list that includes the computer tool (as it is disabled by default):

gptme -t computer   # and whichever other tools you want

You also want to set a screen resolution that is suitable for the vision model you are using.