GitHub Copilot
GitHub Copilot is GitHub's AI tool that helps with writing code and development work.
Overview
GitHub Copilot is GitHub's AI tool that helps improve productivity while working with code. Its official documentation is organized into several feature areas, such as agents, automation and integration with external tools.
Copilot automatically picks a model for each task and can run cloud agents on a set schedule or in response to repository events. You can hand specific tasks to skills or build custom agents, and agent apps made by partners can also be used within GitHub workflows. Tasks that no other tool can handle are automated with the Computer use feature, which operates desktop applications directly.
It suits developers and teams who manage code on GitHub and want to hand repetitive tasks to agents. When Copilot works with code, tools, the file system or the network, you can have it run safely in an isolated sandbox, either locally or in the cloud.
What it is used for
- Get AI help when writing and refining code.
- Run cloud agents automatically on a schedule or when repository events occur.
- Use partner-built agent apps or your own custom agents in GitHub workflows.
- Automate tasks that lack a dedicated tool by operating desktop applications.
- Check and manage the progress of multiple agent sessions on a single page.
Pricing and license
See the official website
Getting started
- Set up Copilot with your GitHub account.
- Follow the Quickstart in the official documentation to turn on the basic features.
- If needed, set up additional agent skills, automation and custom agents.
Links
Related terms and companies
Recent mentions in briefings
- Gemini 4 Argon is now generally available in GitHub Copilot 2026-10-14
- GitHub updates Copilot for JetBrains with default model and MCP server controls and improved chat navigation 2026-10-11
- Claude Haiku 5.5 is now generally available in GitHub Copilot, with token pricing disclosed 2026-10-08
- GitHub makes Copilot local sandboxing generally available, adds local model discovery in the CLI and a dedicated leaked-secret detection model 2026-10-07
Sources
Compiled from public sources (official pages, READMEs). Pricing and features change, so check the official site.