Copilot is the AI coding tool that's quietly become the default for teams already on GitHub. It doesn't try to reinvent your editor — it slides into the one you have and gets out of the way.
What it does
- Inline completions — the bread and butter. As you type, Copilot suggests the next chunk.
- Chat — sidebar conversation with context from your file/workspace.
- Workspace (the agent layer) — multi-step tasks, plans, edits across files. Newer, still maturing.
- Pull request reviews — Copilot can author summaries, suggest changes, even act as a reviewer.
Pricing
- $10/mo individual
- $19/seat/mo business (with team policy controls and no-training defaults)
- $39/seat/mo enterprise (with audit logs, SSO, custom models)
For solo devs the individual plan is hard to beat on price.
Where it wins
- Frictionless rollout in GitHub-centric orgs — same login, same billing, IT already approved.
- Editor breadth — works in pretty much every major editor.
- Enterprise governance — strong policy, audit, and data-handling story.
Where Cursor or Claude Code beat it
- Agent quality — Cursor agent mode and Claude Code still feel a step ahead for autonomous work.
- Editor cohesion — Copilot is a layer; Cursor is an editor. The cohesion shows on long agent runs.
Who should use it
Teams already on GitHub Enterprise, individual devs who don't want to switch editors, anyone who values stability over the bleeding edge.
Frequently asked questions
Copilot vs Cursor — which should I pick?
Copilot if you want AI completions inside the editor you already use. Cursor if you want an AI-native editor where the agent is a first-class citizen. Different categories now.
Is Copilot free for students?
Yes — GitHub Education provides free Copilot Pro for verified students and teachers.
Does my code get used for training?
By default, Business and Enterprise plans don't allow training on your prompts/suggestions. Individual plan has an opt-out toggle.
