Cursor

Cursor is an AI-native code editor that functions as a coding agent, integrating with your terminal, Slack, and GitHub to automate development tasks and boost productivity.

What is Cursor?

Cursor is a code editor built from the ground up for AI-assisted development. It takes your codebase as input and produces code completions, edits, and entire features automatically. It runs on macOS, Windows, Linux, and iOS, and is developed by Anysphere, an applied research team focused on the future of software development.

Key Features

• Composer 2.5 & Agent mode — Write natural language instructions to generate, edit, or refactor code across multiple files, with autonomous agents that can run tests and deploy.
• Cloud agents — Agents run on their own remote computers, allowing them to build, test, and demo features end-to-end without blocking your local machine.
• Codebase indexing — Cursor analyzes your entire codebase for contextual understanding, enabling accurate semantic search and intelligent suggestions across any scale.
• Multi-model support — Choose from models by OpenAI, Anthropic, Google Gemini, xAI, and Cursor’s own models directly within the editor.
• Slack & GitHub integration — Collaborate via Slack (ask Cursor to take a stab at features) and get PR reviews acted on automatically through the agent.
• Automations — Schedule recurring agents to handle maintenance, security fixes, or deployments on a trigger.
• Cross-platform availability — Native desktop apps for macOS, Windows, and Linux, plus a mobile app for iOS.
• CLI tool — Use Cursor’s agent directly from the command line for terminal-centric workflows.

Who is it for?

• Individual developers — Accelerate daily coding tasks, from quick completions to building full features with agent instructions.
• Engineering teams — Share a consistent AI assistant that understands the entire codebase, integrates with Slack and GitHub, and works on any machine.
• Enterprise organizations — Trusted by over half of the Fortune 500 (including Stripe and NVIDIA) for secure, scalable development.
• Researchers — Leverage state-of-the-art models for experiment-driven development, with published research on RL, multi-agent collaboration, and semantic search.

What can you do with Cursor?

• Build features end-to-end — Ask the agent to create a landing page from design docs; it will write code, preview, and iterate until done.
• Analyze usage patterns — Have the agent query workspace exports and generate analytics on how your team uses Tab vs Agent modes.
• Create mission control interfaces — Build complex UIs like an expose-style window manager using natural language instructions across files.
• Automate repetitive work — Set up always-on agents to handle PR reviews, dependency updates, or live dashboards on a schedule.

How does Cursor work?

You write natural language instructions in a composer pane or via the CLI. Cursor’s agent reads your codebase (indexed locally or via cloud), plans the changes, and executes them across files. It can run shell commands, open browser previews, and show you a diff before applying any change.

FAQ

Is Cursor free?

Cursor offers a free tier with limited completions and agent usage, plus paid Pro and Business plans for higher limits and team features. Exact pricing is available on the website.

What platforms does Cursor support?

Cursor provides native installers for macOS (Intel and Apple Silicon), Windows (x64 and Arm64), Linux (deb and rpm for x64), and an iOS mobile app. The CLI works on any platform.

Which AI models can I use?

Cursor supports models from OpenAI (GPT-4o, GPT-5.5), Anthropic (Claude Opus 4.8, Composer 2.5), Google Gemini 3.1 Pro, xAI Grok 4.3, and its own fine-tuned models. You can switch per task.

Can Cursor understand my entire codebase?

Yes. Cursor indexes your entire project for context-aware completions, semantic search, and intelligent navigation. It supports any programming language and scales to large monorepos.

How do I get started?

Download the desktop app for your platform from cursor.com, sign in, and open a project. Start typing or press Cmd+K to compose instructions. The agent will guide you through your first task.