GitHub Copilot

What is GitHub Copilot?

GitHub Copilot is an AI coding assistant built by GitHub, which Microsoft owns, and powered by OpenAI models. Its original trick was completion: start typing a function and Copilot suggests the rest of it, often across many lines, using the surrounding code and comments as context. For millions of developers it turned into an everyday pair programmer.

Since launch it has grown far beyond autocomplete. Copilot Chat answers questions about your codebase, Copilot agent mode plans and executes multi-step tasks, and Copilot pull requests can even write summaries and fix review comments. There is also a CLI so the assistant follows you into the terminal.

The tool launched as a technical preview in June 2021 and reached general availability in 2022. In late 2024 GitHub introduced a free tier, making the basic assistant available to everyone with a GitHub account, which cemented its role as the default AI assistant for many teams.

Key Features

• Inline code completion that suggests multi-line edits
• Chat panel grounded in your repository context
• Agent mode for multi-file, multi-step tasks
• Pull request summaries and review support
• Works across VS Code, JetBrains, Neovim, Xcode and the CLI
• Copilot Workspace for planning features from GitHub issues
• Custom instructions and organization policy controls

How to get started

Install the Copilot extension in your editor of choice, sign in with your GitHub account, and enable suggestions. From there you just write code as usual and accept or reject the grayed-out suggestions with Tab. To go further, open the chat panel, ask questions about your repo, and try agent mode on a small task to see how it plans changes.

• Install the GitHub Copilot extension in your editor
• Sign in with your GitHub account
• Start typing and accept suggestions with Tab
• Ask questions in Copilot Chat
• Try agent mode on a small refactoring task

Use cases

• Faster everyday coding with inline completions
• Explaining unfamiliar code and frameworks via chat
• Generating unit tests and documentation
• Summarizing and reviewing pull requests
• Drafting features directly from GitHub issues
• Learning a new language by example, in context

Pricing and licensing

GitHub Copilot moved to a freemium model: the free tier gives everyone a limited number of completions and chat requests per month, while Pro, Business and Enterprise plans add unlimited use and stronger features. It is available to individuals and organizations through GitHub plans or as a standalone subscription.

There were early questions about whether AI suggestions could copy licensed code, and GitHub added a code reference filter for paid plans to help with that. The current pricing details and included models change regularly, so the official plans page is the source of truth.

Pros and cons

Copilot's strengths are reach and polish. It works in every major editor, the completion quality keeps improving, and because it is built into GitHub, the pull request and issue integrations feel seamless. The free tier also lowers the barrier to trying it.

The trade-offs are model lock-in and noise. You get the models GitHub provides rather than choosing freely, and suggestions are only as good as the context around them, so messy codebases produce messier help. Teams with strict IP policies may also prefer self-hosted or on-device assistants.

Alternatives

• Cursor: AI-first editor with deep codebase awareness
• Claude Code: Anthropic's terminal coding agent
• Windsurf: agentic AI editor built on a VS Code fork
• Continue: open-source AI assistant for VS Code and JetBrains
• Tabnine: privacy-focused completion for enterprises