Cursor logo

Cursor

by Anysphere

AI CodingAI CodingAI Code EditorAI Coding AgentAI Developer ToolAI AssistantAI AgentCode Generation

AI-powered code editor and coding agent for building, editing, debugging, reviewing, and shipping software.

Cursor is an AI-powered code editor and coding agent developed by Anysphere for software development. It combines an AI-native editor with codebase understanding, intelligent code completion, Agent, Plan and Ask modes, terminal tools, code editing, web search, MCP integrations, cloud agents, code review, and multiple frontier AI models. Developers can use Cursor to understand existing repositories, plan and implement features, fix bugs, review changes, automate development tasks, and work across local, cloud, and terminal environments. Cursor supports models from providers including OpenAI, Anthropic, Google, xAI, and its own Cursor models.

Provider

Anysphere

Category

AI Coding

Available On

Windows, macOS, Linux, Web, iOS, CLI

Launched

Mar 2023

Use Cases

Software Development, Coding, Code Refactoring, Bug Fixing

Best For

Developers who want AI-assisted coding, repository understanding, autonomous agents, code review, debugging, and software development automation

🧠 Models (7)

Grok 4.7v4.7

Reasoning

Context Window

Long context

Input

text, code

Output

text, code

Grok 4.6v4.6

Reasoning

Context Window

Long context

Input

text, code

Output

text, code

Grok 4.5v4.5

Reasoning

Context Window

Long context

Input

text, code

Output

text, code

Composer 2.5v2.5

Coding Agent

API

Context Window

Long context

Input

text, code, files

Output

text, code

GPT-5.6 Lunav5.6 Luna

Reasoning

Context Window

Model dependent

Input

text, code

Output

text, code

Claude SonnetvCurrent

Coding and Reasoning

Context Window

Model dependent

Input

text, code

Output

text, code

GeminivCurrent

Multimodal

Context Window

Model dependent

Input

text, code, image

Output

text, code

✨ Features (15)

Cursor Agent

Autonomous coding agent that can understand a codebase, edit files, run commands, search the web, and complete complex development tasks

Plan Mode

Creates and refines an implementation plan before making code changes

Ask Mode

Provides read-only codebase exploration and answers without modifying files

Tab Completion

Provides AI-powered code completions and predicts useful edits based on surrounding code and recent changes

Codebase Understanding

Indexes and searches project code to provide relevant context to AI models

Multi-File Editing

Can make coordinated changes across multiple files in a repository

Cloud Agents

Run coding agents in isolated cloud environments that can continue working while the developer is away

Cursor Projects

Coordinate larger development projects by delegating work to multiple agents and maintaining shared context

MCP Support

Connect external tools, services, and data sources through the Model Context Protocol

Cursor CLI

Use Cursor coding agents directly from a terminal for interactive development, automation, and CI workflows

Bugbot

AI-powered code review tool that analyzes changes and helps identify potential issues

Cursor Rules

Define project-specific instructions and coding guidelines for Cursor agents

Model Selection

Choose from Cursor models and supported frontier models from multiple AI providers

Cursor Router

Automatically selects models based on optimization modes such as cost, balance, and intelligence

Privacy Mode

Provides a privacy setting designed to prevent code from being used for model training

💰 Pricing

Free Plan

Free Trial

Paid Plan

Enterprise

Starting Price

$20/month

Billed monthly

API Pricing

Cursor models and third-party models use usage-based token pricing; additional usage can be billed at published model rates.

💬 Prompt Library (5)

Fix Bug

Coding

Find and fix bugs across a codebase

Analyze this project and find the root cause of the following bug: [DESCRIBE BUG]. Inspect the relevant files, explain the cause, make the required fix, and verify that the fix does not introduce new issues.

Build Feature

Coding

Build a complete feature inside an existing project

Implement this feature: [FEATURE DESCRIPTION]. First inspect the existing project architecture and identify the files that need changes. Then create a clear implementation plan, make the changes, and test the result.

Code Review

Coding

Perform an AI-assisted code review

Review the current codebase for security, performance, maintainability, architecture, error handling, and potential bugs. Identify the most important issues and suggest specific fixes.

Refactor Code

Coding

Improve existing code while preserving behavior

Refactor the following part of the project without changing its intended behavior. Improve readability, maintainability, performance, and code structure. Explain the changes and verify the result with appropriate tests.

Understand Project

General

Understand an unfamiliar codebase before making changes

Analyze this repository and explain its architecture, main components, data flow, important files, dependencies, entry points, and how the major features work. Do not modify any files.

🔗 Integrations (8)

🎯 Related AI Coding Tools

View all →

⚠️ Limitations

  • Advanced Agent and model usage is subject to plan-specific limits
  • Different AI models have different usage rates and context capabilities
  • Cloud Agents require repository access and temporarily store repository copies while agents work
  • AI-generated code can contain bugs, security vulnerabilities, or incorrect implementations
  • Some models and features are restricted to specific subscription plans
  • Third-party model requests may be subject to the selected model provider's availability and pricing
  • Cursor requires internet connectivity for many AI-powered features
  • AI agents can make unintended changes, so generated changes should be reviewed and version-controlled

📚 Sources