Cursor
An AI-first code editor built on VS Code with deep in-editor agent features, parallel agents, in-house Composer models, and a plugin marketplace.
Cursor is an AI-first code editor forked from VS Code, so extensions, themes, and keybindings carry over. It layers on tab completion, inline natural-language edits, an agent mode that reads, writes, and runs commands across files, and parallel agents across repos and worktrees. Freemium: a free Hobby tier, with Pro and Teams raising included usage.
Cursor is a code editor forked from VS Code that puts AI assistance at the center of the editing experience. Because it is built on the VS Code codebase, existing extensions, themes, keybindings, and settings carry over, so the learning curve is mostly about the AI features layered on top.
It is aimed at developers who want inline completions and chat-driven edits without leaving the editor or copy-pasting between a browser and their codebase. Cursor indexes your project so the model can reference relevant files when answering or editing.
Highlights
- Tab completion — multi-line, context-aware suggestions that can edit across the current file.
- Inline edits — select code, press the edit shortcut, and describe the change in natural language.
- Agent mode — a chat agent that can read, write, and run commands across multiple files to complete a task.
- Parallel agents — Cursor 3.0 (April 2026) rebuilt the interface agent-first: run many agents at once across repos — locally, in git worktrees, in the cloud, or over SSH — with side-by-side agent tabs.
- Codebase context — reference files, symbols, or docs with
@mentions so the model grounds its answers in your code. - Model choice — switch between frontier models (Anthropic, OpenAI, and others) per request, including Cursor's in-house Composer models tuned for fast agentic coding.
- Plugin marketplace — reviewed plugins (Atlassian, Datadog, GitLab, and more) extend the editor and its agents.
In an AI-assisted workflow
Cursor fits where you already write code. A common loop is to describe a change in the chat panel, let agent mode draft edits across files, then review the diff before accepting. You can scope context explicitly:
@components/Button.tsx Refactor this to accept a `variant` prop
and update all call sites in @app/.NOTE
Cursor reviews and applies edits as inline diffs you accept or reject, so the AI never silently overwrites your files.
Good to know
SpaceX completed its acquisition of Cursor in August 2026, and Cursor's plans now include Grok models and Grok Bot alongside its in-house Composer models. Cursor is available on macOS, Windows, and Linux. The free Hobby tier includes limited AI usage; paid Individual (Pro and up) and Teams plans raise included usage and unlock premium models, with on-demand usage billed beyond the included amount. You can also supply your own API keys. Because it is a separate application rather than an extension, it runs alongside (not inside) a standard VS Code install.
Frequently asked questions
- What is Cursor?
- Cursor is a code editor forked from VS Code that puts AI at the center of the editing experience: multi-line tab completion, inline edits described in natural language, and an agent mode that can read, write, and run commands across multiple files. Cursor 3.0 rebuilt the interface agent-first, running many agents at once across repos — locally, in git worktrees, in the cloud, or over SSH.
- How much does Cursor cost?
- Cursor is freemium. The free Hobby tier includes limited AI usage; paid Individual (Pro and up) and Teams plans raise included usage and unlock premium models, with on-demand usage billed beyond the included amount. You can also supply your own API keys.
- Cursor vs VS Code?
- Cursor is a separate application built on the VS Code codebase, not an extension — it runs alongside (not inside) a standard VS Code install. Existing extensions, themes, keybindings, and settings carry over, so the learning curve is mostly the AI features layered on top.
Filed under
ide · editor
Related
- Devin Desktop (formerly Windsurf)An agentic IDE — formerly Windsurf, now Devin Desktop from Cognition AI — with flows that take multi-step actions across your codebase.
- ContinueAn open-source IDE extension for building custom AI coding assistants.
- Claude CodeAnthropic’s official agentic coding tool that runs in the terminal, IDE, and web.
- How to Test AI-Generated CodeAI writes the code; tests decide whether to trust it. The verification stack for agent-written changes — contracts, generated tests, and the review that's left.
- Vibe CodingVibe coding is building software by describing intent in natural language and letting an AI agent write the code, judging results by behavior.
- Vibe Coding in 2026: What It Is, When It Works, When It BitesAn honest guide to vibe coding — where prompt-and-accept development genuinely pays, where it accumulates risk, and the guardrails that make it professional.
- Cursor vs Claude Code vs GitHub Copilot vs Windsurf in 2026A practical, opinionated comparison of the four mainstream AI coding tools — form factor, agentic depth, model choice, and who each one is for.
- ZedA high-performance, multiplayer code editor with built-in AI assistance.