Back
GGitHub Copilot logo

GitHub Copilot

An AI coding assistant in editors and GitHub for completion, chat, and review.

AI assistantsAI codingChat assistantsCode assistantsFree plan
Official website

Plans and pricing

Official pricing verified
Access
Free plan
Paid plans from
$10/mo
Billing
Monthly subscription

Free

$0

Pro

$10/mo

Pro+

$39/mo

Business

$19/user/mo

Official pricing

DevPrice organizes public information and does not sell GitHub Copilot subscriptions. Prices and availability are determined by GitHub Copilot.

What is GitHub Copilot?

GitHub Copilot is an AI coding product that spans code editors and the GitHub collaboration workflow. Originally centered on inline completion, it now includes question answering, implementation planning, executable agents, cloud coding tasks, and code review, allowing AI to participate from writing code through preparing and reviewing a pull request.

Inside an editor, completion follows the current input, Ask mode explains or discusses implementation, Plan mode prepares a reviewable approach, and Agent mode can identify relevant files, propose edits, and run terminal commands. Developers can choose between receiving a suggestion and delegating a more complete change according to the scope and risk of the task.

Within a GitHub repository, Copilot cloud agent can take an issue or prompt, study the code in an isolated environment, make changes, and open a pull request. Copilot code review examines existing changes and produces suggestions that can be applied. The product fits teams already organized around GitHub, pull requests, and mainstream IDEs, but model allowances, agents, review features, and organization policies vary by plan, and generated code still requires human verification.

Key features of GitHub Copilot

  • Provide inline completion in the editor

    Copilot suggests single-line or multi-line code from the current file, cursor position, and surrounding context. Developers can accept, ignore, or revise the suggestion without leaving the file.

  • Match task scope with Ask, Plan, and Agent

    Ask is suited to explanations and suggestions, Plan prepares detailed implementation steps, and Agent handles multi-step work that requires finding files, editing code, running commands, and iterating.

  • Delegate an issue to a cloud agent

    Copilot cloud agent can receive a coding task on GitHub, work in an isolated environment, and open a pull request. Users can inspect the session and request further changes through comments.

  • Review pull requests and suggest fixes

    Copilot code review examines changes with repository context and leaves suggestions on a pull request. Some fixes can be applied directly or passed to the cloud agent for a follow-up branch.

  • Work across GitHub and multiple development environments

    Beyond GitHub.com, Copilot integrates with environments including VS Code, Visual Studio, JetBrains IDEs, Xcode, the command line, and mobile. Available features vary by client and organization policy.

  • Ground work in repository conventions

    Repositories can provide custom instructions, AGENTS.md, agent skills, and MCP context so Copilot understands coding standards, test requirements, and external systems instead of relying only on generic guidance.