AI Coding tool profile
Cursor
An AI-first code editor (VS Code fork) with autocomplete, inline chat, and agent mode that can plan and apply multi-file edits across a project.
Independent research from official sources · Sources · How we review
- Best for
- Developers who want AI assistance inside a familiar editor, from fast autocomplete to agent-style multi-file changes, with a clear free tier to start.
- Starting point
- Free Hobby plan plus paid tiers: Pro from $20/month, Pro+ $60/month, Ultra $200/month, and Teams from $40...
- Free plan
- Yes — Hobby plan is free with no credit card required and includes limited Agent requests plus access to Composer.
- Key limitation
- Pricing is credit-based beyond the free tier; manual frontier-model selection consumes credits faster than Auto mode.
Choose it if
Strongest for
Built for
Avoid it if
- Agent behavior, model availability, and credit costs can change after the review date.
- JS-rendered pricing page: confirm current plan limits and credit prices on the official site.
Free-plan suitability
Hobby plan is free with no credit card required and includes limited Agent requests plus access to Composer; paid tiers (Pro $20/month, Pro+, Ultra, Teams $40/user/month) add extended Agent limits, frontier models, cloud agents, and usage-based billing.
Pricing
Free Hobby plan plus paid tiers: Pro from $20/month, Pro+ $60/month, Ultra $200/month, and Teams from $40 per user per month, with credit-based model usage. Check the official site for current plan boundaries and credit costs.
Verify the current plan, credits, output limits, and commercial terms before purchasing.
What Cursor does
- AI code editor (IDE) with agentic development: Cmd+K targeted edits up to a full-autonomy agent mode that plans, searches, and builds
- Cloud agents work autonomously on their own computers to build, test, and demo features end to end for review
- Runs across surfaces: desktop IDE, CLI in the terminal, Slack collaboration, GitHub PR review with Bugbot, and an iOS mobile app
- Automations: always-on agents that run on schedules or triggers to build, maintain, and fix software (e.g., fix CI failures on main)
- Composer plus frontier model access (e.g., Grok 4.6, Opus 5) with MCPs, skills, hooks, and a marketplace for extensions
- Secure codebase indexing, semantic search, and context-aware completions; usage-based pricing with on-demand overage after included model limits; SOC 2 certified
- Remote control for local agents (Oct 6, 2026): see and reply to agents running on your computer from the Cursor iOS app; on by default except for Enterprise organizations, where admins enable it, and the computer must stay on, online and paired by approving a request in the desktop app (Cursor changelog, checked Oct 13, 2026)
Is it legit?
Who is behind Cursor, where it is based, and what is publicly known about its funding. Company details change — verify on the official site before committing.
- Company
- Anysphere (Cursor)
Evaluation guide
Step-by-step evaluation guides for Cursor — workflow fit, pricing, rights, and a same-input test before paying.
Compare Cursor with alternatives
Side-by-side pages that weigh Cursor against overlapping tools on the same five fields.
Official sources
These links are retained beside the catalog record so changing details can be rechecked.