GitHub Copilot review

What it is

GitHub Copilot is the assistant that created this category and remains its default. It provides inline completions, a chat sidebar that understands your open files, and an agent mode that can plan and apply multi-file changes and open a pull request.

Where it shines

Ubiquity. It runs in VS Code, JetBrains, Neovim, Visual Studio and Xcode, so a team can adopt it without anyone changing editors — a bigger deal than it sounds, because editor migration is where most coding-tool rollouts die.

Completions remain excellent for the everyday middle of programming: boilerplate, tests, repetitive transformations, the second half of a function whose first half you already wrote. Model choice matters too — you can point Copilot at different frontier models depending on the task rather than being locked to one vendor's roadmap.

For enterprises the decisive factor is procurement. It arrives through an existing Microsoft relationship with policy controls, audit logs and IP indemnification already in place, which turns a six-month security review into a purchase order.

Where it falls short

Depth of context. Competitors that index the whole repository answer architectural questions Copilot cannot, and its agent mode, while improving, still trails Cursor and Claude Code on large refactors — it is more likely to make a plausible local change that breaks something three files away.

And the perennial caveat: suggestions look authoritative regardless of correctness. Junior developers accepting completions uncritically is a real code-quality risk, which is a management problem more than a tooling one.

Pricing in plain English

Free tier with 2,000 completions monthly, and free for students and open source maintainers. $10 a month for Pro. $39 for Pro+ with the largest premium request allowance. $19 per user for Business with governance and indemnity.

Who should use it

Almost every professional developer, as a baseline. Add Cursor or Claude Code on top when you need heavy multi-file work.