Cursor

Freemium

By AI Bookmarks Editorial Team, Directory editors Last reviewed 2026-10-23

Visit Cursor Official site · opens in a new tab

Code editor built around AI agents that read the repository, edit several files and run commands.

Quick facts

Pricing	Freemium
Free tier	Yes — free tier with a limited number of agent requests
Platforms	macOS, Windows, Linux
Developer	Anysphere
Website	https://cursor.so

What Cursor is

Cursor is a code editor built on the same foundation as VS Code, with AI woven through it rather than added as an extension. It offers inline completion, a chat panel that can see the repository, and an agent mode that plans a change, edits several files, runs commands and reports back. Because it inherits the VS Code extension ecosystem, switching from that editor is mostly a settings import. The distinguishing feature is codebase awareness: the project is indexed so answers reference real files rather than whatever was pasted. The limits are cost and control: usage is metered by requests to the stronger models, long agent sessions consume the allowance quickly, and generated changes need review like any other contribution. Individual and team plans differ more in model access than in editor features. Rules files let a team encode conventions that the agent follows across tasks.

Best for

Multi-file changes inside an existing repository
Teams already familiar with VS Code
Reviewing an agent diff rather than typing every line

Not ideal for

Work that must stay entirely offline
Editors where every change needs manual control

Pros

Codebase indexing makes answers reference real files
Inherits the VS Code extension ecosystem
Agent mode edits several files in one task

Cons

Strong-model requests are metered and run out
Agent output needs the same review as any patch
Smaller vendor with a shorter track record than Microsoft

Pricing in detail

The free tier includes a limited number of agent requests and completions. Paid plans are billed monthly or annually and raise the request allowance and unlock stronger models; team plans add administration and pooled usage. Because the meter is model requests rather than seats, a heavy refactoring day consumes a plan faster than a week of small edits. Check whether unused requests roll over before choosing a tier.