GitHub Copilot
What you can do
Your AI coding agent
Start from an issue or pull request
Open a Copilot session with the task, discussion, and repository context already connected.
Run tasks in parallel
Give each task its own session and track progress from one view.
Continue work across surfaces
Let long-running tasks continue, then check progress or resume them from the Copilot app, CLI, editor, github.com, or GitHub Mobile.
Automate recurring work
Run reusable agent tasks on demand, on a schedule, or from supported triggers.
Review local changes
Find code quality and security issues before opening a pull request, then ask Copilot to address the findings you choose.
Review pull requests
Combine Copilot code review with team feedback and checks. Delegate follow-up changes to the Copilot cloud agent when needed.
Models
Pick the model that fits the job
Let Copilot choose
Match each task with one or more models, balancing quality, speed, and cost.
Choose your model
Select a specific model when you want direct control.
Add cloud or local models
Connect supported cloud model providers with your own API key, or run models locally with Ollama and Foundry Local.
Setup
Make Copilot yours
Start with your repository
Use code, pull requests, issues, and search results to give Copilot context for the task.
Add your instructions
Define coding standards and repeatable workflows with repository instructions and skills.
Connect your services
Use Copilot Connectors to bring in data and tools from the services you already use.
Find more tools
Discover MCP servers, skills, and plugins on GitHub, then add the ones you choose.
Resources
Go further with Copilot
Docs
Get hands-on training with the Copilot App
Use cases
See Copilot in action
Changelog
Explore the latest releases
Frequently asked questions
