What is Self-Operating Computer used for?

Self-Operating Computer is an MIT-licensed Python framework that lets multimodal models view a screen and control the mouse and keyboard to complete an objective. It is a good fit for experimenting with computer-use AI agents and automating desktop tasks without scripts.

Is Self-Operating Computer open source?

Yes. Self-Operating Computer is open source under the MIT licence. Its source code is on GitHub at OthersideAI/self-operating-computer and is written mainly in Python.

Is Self-Operating Computer free?

Yes. Self-Operating Computer is open source, so the software itself is free to use.

What are some alternatives to Self-Operating Computer?

Similar open-source tools in the AI Tools category include Nanobot, GPT Researcher and Goose. SaaS products in the same category include Aera Browser, August and CalmPilot.

Is Self-Operating Computer actively maintained?

The most recent commit to Self-Operating Computer was on 19 September 2025, and the latest release is v1.5.8, published on 28 February 2025. The project has 10k stars on GitHub.

Self-Operating Computer: open-source AI tools

About Self-Operating Computer

Self-Operating Computer is a Python framework that lets a multimodal AI model operate a computer the way a person would. The model receives screenshots of the display and decides on mouse and keyboard actions to reach a goal the user states. The README describes it as an early example of full computer use, first released in November 2023.

It is designed to work with different models and currently integrates with GPT-4o, GPT-4.1, o1, Gemini Pro Vision, Claude 3, Qwen-VL and LLaVA. The tool is started with the operate command, a mode flag selects the model, and OpenAI's gpt-4o is the default. On macOS the terminal app must be granted Screen Recording and Accessibility permissions, and an API key is supplied for the chosen provider.

The project is released under the MIT license and runs locally on your machine, using PyAutoGUI for input automation according to its topics. Because the agent controls the real desktop, it is best run in a controlled environment and with supervision.

Key features

Model views screen and issues mouse and keyboard actions
Supports GPT-4o, Gemini, Claude 3, Qwen-VL and LLaVA
Command-line operate entry point
Model selection through a command-line mode flag
Objective-driven task execution

Good fit for

→Experimenting with computer-use AI agents
→Automating desktop tasks without scripts

Built with: Python
Tags: computer-use; ai-agent; automation; multimodal; python; pyautogui; gpt-4o; llm

Self-Operating Computer: questions and answers

What is Self-Operating Computer used for?: Self-Operating Computer is an MIT-licensed Python framework that lets multimodal models view a screen and control the mouse and keyboard to complete an objective. It is a good fit for experimenting with computer-use AI agents and automating desktop tasks without scripts.
Is Self-Operating Computer open source?: Yes. Self-Operating Computer is open source under the MIT licence. Its source code is on GitHub at OthersideAI/self-operating-computer and is written mainly in Python.
Is Self-Operating Computer free?: Yes. Self-Operating Computer is open source, so the software itself is free to use.
What are some alternatives to Self-Operating Computer?: Similar open-source tools in the AI Tools category include Nanobot, GPT Researcher and Goose. SaaS products in the same category include Aera Browser, August and CalmPilot.
Is Self-Operating Computer actively maintained?: The most recent commit to Self-Operating Computer was on 19 September 2025, and the latest release is v1.5.8, published on 28 February 2025. The project has 10k stars on GitHub.