GitHub Copilot

Editor's Review

GitHub Copilot in 2026 is a $10/month completion engine with an identity crisis. The base prices didn't move — Free (limited), Pro $10/month, Pro+ $39, Max $100 (new), Business $19/user, Enterprise $39/user — but on June 1, 2026 the billing engine switched from flat "premium requests" to metered GitHub AI Credits billed per token at each model's published API rate. Completions and next-edit suggestions stay unlimited on paid plans; chat, CLI, agents, code review, and Workspace all draw from the credit pool, with overage at $0.01/credit and hard spending caps added in July. Why the change? A handful of agentic users were costing more than their plan price. And the market noticed: Copilot's share fell in the Stack Overflow 2026 survey (67% → 51%) while Cursor and Claude Code gained. Architecturally it's still a completion engine with agent mode bolted on — agent mode does partial multi-file refactors then asks you to finish by hand, trailing both rivals on cross-file autonomy. Only 9% of senior devs prefer it versus 46% for Claude Code (JetBrains 2026). Who should pay? Mixed-IDE teams that need one license across VS Code, JetBrains, Neovim, and Xcode with GitHub PR integration — nobody else covers that footprint at $10/month, and unlimited completions remain genuinely good. Who shouldn't? Anyone wanting a real agent: Cursor ($20/month, better Composer agent) or Claude Code ($20/month entry, terminal-native) are both more autonomous. Copilot is the best-value autocomplete in the business; it's just no longer the agent conversation.