Cursor
by Anysphere, Inc. (SpaceX)
The AI-native code editor that reads your whole repository and edits across it.
Cursor is an AI-native code editor built on VS Code by Anysphere, adding agents that edit across multiple files, run commands, review pull requests and work from codebase-wide semantic retrieval. Engineering organisations adopt it as a governed rollout with SSO, SCIM, audit logs and privacy mode rather than as an individual developer plugin.
Cursor is an AI-native code editor built as a fork of VS Code by Anysphere, extending the familiar editor with agentic capabilities: multi-file edits, a background agent that runs tasks asynchronously, codebase-wide semantic retrieval, Bugbot for automated code review, and support for Model Context Protocol servers, custom rules, skills and hooks. Its core design bet is that AI assistance belongs in the editor's execution path rather than in a sidebar chat — the agent reads the repository, proposes multi-file diffs, runs commands and iterates on test output. Cursor is model-agnostic, routing to frontier models from multiple providers alongside its own Composer model. Enterprise adoption has been the company's real engine: by mid-2026 roughly $2.6 billion of its approximately $4 billion annualised revenue came from business customers, with more than half the Fortune 500 reported as users. Team and enterprise tiers add SAML/OIDC SSO, SCIM seat management, pooled usage across an organisation, repository and model access controls, audit logs, service accounts, an AI code-tracking API and team-wide privacy mode. Cursor states a SOC 2 Type II report is available on request, commits to at-least-annual third-party penetration testing, and offers Privacy Mode on free and paid plans, which contractually and technically prevents training on customer code, alongside customer-managed encryption keys. Its trajectory is unusual even by 2026 standards: revenue doubled roughly every three months from $100 million ARR in January 2025, and on 16 June 2026 SpaceX agreed to acquire Anysphere for $60 billion in stock — the largest venture-backed startup acquisition on record — closing 14 August 2026 and folding Cursor into a new SpaceXAI division.
VPs of Engineering rolling AI coding out across a whole org who need SSO, SCIM, audit logs, model access controls and a defensible answer on whether code trains a vendor model.
Agents that edit across the real repository and run the real commands, under org-level identity, policy and usage controls rather than per-developer subscriptions.
At a Glance
- Category
- Developer Tools
- Pricing
- Freemium, Subscription, Usage-based
- Target Market
- CTOs, VPs of Engineering, Enterprise Developers, Platform Engineers, Engineering Managers
- Deployment
- Cloud-first
- Founded
- 2022
- Headquarters
- San Francisco, United States
- Customers
- Reported at more than half of the Fortune 500 (press reports, mid-2026)
Key Features
- ✓Multi-file agentic editing
The agent reads the repository, proposes diffs spanning many files, runs commands and iterates on the resulting test output.
- ✓Background and cloud agents
Tasks run asynchronously away from the editor session, so long refactors proceed without blocking interactive work.
- ✓Codebase-wide semantic retrieval
Indexes the whole repository so the model reasons about actual project code rather than only the open file.
- ✓Bugbot automated code review
Reviews pull requests agentically as a first pass, catching mechanical defects before they consume human reviewer attention.
- ✓MCP, rules, skills and hooks
Model Context Protocol servers and shareable rules let teams standardise how agents access internal tools and conventions.
- ✓Enterprise identity and governance
SAML/OIDC SSO, SCIM seat management, audit logs, service accounts, repository and model access controls, and pooled organisation-wide usage.
- ✓Privacy Mode and CMEK
Available on free and paid plans, it prevents training on customer code with both technical and contractual protection, plus customer-managed encryption keys.
Capabilities
Use Cases
- •Large multi-file refactor or framework migration
An agent applies a consistent change across hundreds of files and iterates until the test suite passes.
- •First-pass automated pull request review
Bugbot flags mechanical defects before human reviewers engage, shortening review cycles on high-volume repositories.
- •Governed org-wide AI coding rollout
Platform teams enforce model access, repository scope and privacy mode centrally instead of auditing individual developer accounts.
- •Onboarding into an unfamiliar codebase
New engineers query repository-wide semantic context instead of interrupting colleagues to trace how a subsystem works.
- •Internal tool access through MCP
Standardised MCP servers let agents call internal APIs and services without each developer wiring their own integration.
Ideal For
Best For
- ✓Org-wide AI coding rollouts that need SSO, SCIM, audit logs and pooled billing rather than individual seats
- ✓Multi-file refactors and migrations across an existing codebase where single-file completion is not enough
- ✓Automated pull request review via Bugbot as a first pass before human reviewers
- ✓Teams standardising agent tooling through MCP servers, shared rules and skills across the organisation
- ✓Migration from VS Code where preserving extensions, keybindings and muscle memory matters
Not Ideal For
- ✗Organisations that require an on-premises or air-gapped deployment — Cursor is a cloud-connected product and the security page describes no self-hosted option
- ✗Teams needing regulated compliance beyond SOC 2 Type II: ISO 27001 and HIPAA are not claimed on the security page, so healthcare and similar buyers should confirm before committing
- ✗Anyone who cannot tolerate variable spend — every plan bundles a usage allowance and then bills on-demand model usage in arrears, which is the most common cost surprise
- ✗Security-conservative teams that have not addressed agent execution risk: a published vulnerability class allows a repository to autorun code on a developer's machine when opened
Deployment
Market & Ratings
Reported at more than half of the Fortune 500 (press reports, mid-2026)
Market Analysis
Pros
- ✓Agentic editing across a real repository with command execution is materially more capable than file-scoped completion
- ✓Enterprise controls are genuinely deep — SSO/SCIM, pooled usage, repository and model access controls, audit logs and an AI code tracking API
- ✓Privacy Mode with no-training guarantees is available on every plan, including free, which is rare in this category
- ✓VS Code foundation means near-zero migration cost for teams already there
- ✓Model-agnostic routing across frontier providers plus its own Composer model avoids single-model dependency
Cons
- ✗Usage-based billing in arrears past each plan's included allowance makes real monthly cost hard to predict, which is the most consistent practitioner complaint
- ✗Agent execution has a documented security surface: a reported vulnerability class lets a repository autorun malicious code when opened, so untrusted repositories are a real risk
- ✗Model knowledge lags newer libraries and practitioners report suggestions referencing outdated packages despite web search compensation
- ✗Cloud-connected only — no on-premises or air-gapped option, which rules it out for the most restrictive environments
- ✗The $60 billion SpaceX acquisition, closed 14 August 2026, folds Cursor into a new SpaceXAI division; roadmap, pricing and data-handling continuity under a new owner are open questions for anyone signing a multi-year agreement
Pricing
Hobby
$0
- ✓Limited agent requests
- ✓Access to Composer
- ✓No credit card required
Pro
From $20/mo
- ✓Extended agent limits
- ✓Frontier model access
- ✓MCPs, skills and hooks
- ✓Cloud agents
- ✓Usage-based Bugbot billing
Teams Standard
From $40/user/mo
- ✓Centralised billing and administration
- ✓SAML/OIDC SSO
- ✓Bugbot agentic code review
- ✓Team marketplace for rules and skills
- ✓Usage analytics
- ✓Team-wide privacy mode
Enterprise
Contact for pricing
- ✓Pooled org-wide usage
- ✓SCIM seat management
- ✓Repository and model access controls
- ✓Audit logs and service accounts
- ✓AI code tracking API
- ✓Invoice/PO billing
- ✓Priority support
Published list pricing starts at $0 for Hobby and $20/month for Pro, with Teams at $40 per user per month; higher-usage Pro+ and Ultra individual tiers and a Teams Premium tier exist above these but their prices are not stated on the public plan cards. The critical mechanic is that every plan includes a fixed model-usage allowance and then bills on-demand usage in arrears once that is consumed, so the sticker price is a floor rather than a cap. Enterprise is quote-only and is where pooled organisation-wide usage, SCIM, repository and model access controls, audit logs, service accounts, the AI code tracking API and invoice/PO billing live.
Security & Compliance
Sources
This page was written from 6 sources, 4 on domains other than cursor.com.
Stay Ahead of the Curve
Weekly enterprise AI insights for technology leaders. No spam, no vendor pitches—unsubscribe anytime.
SubscribeRelated Products
Raindrop
AI agent monitoring that catches silent production failures: Sentry for AI agents
GitLab Duo Agent Platform
Agentic AI across the whole GitLab DevSecOps lifecycle: planning, coding, code review, CI/CD and security agents under one governance model
CodeRabbit
AI code review and agentic change management for teams shipping human- and machine-written code
Qodo
Agentic code review and governance layer for teams shipping AI-generated code at scale
Mentioned In
GPT-5.4 vs Claude Opus: Which Actually Saves You Money?
GPT-5.4 vs Claude Opus 4.6: cost analysis, performance benchmarks, and ROI calculations. Find out which saves your team 40% on AI infrastructure spend.
March 7, 2026Enterprise AIAmazon's AI Code Created Outages. Now Humans Must Sign Off.
Enterprise AI analysis: Amazon's AI Code Created Outages. Now Humans Must Sign Off.. Strategic insights, ROI considerations, and implementation guidance for ...
March 11, 2026Code AIGitHub Copilot vs Cursor vs Replit: Enterprise Code AI Battle
Enterprise AI analysis: GitHub Copilot vs Cursor vs Replit. Strategic insights, ROI considerations, and implementation guidance for technical and business le...
March 11, 2026EngineeringThe Real Cost of AI Transformation: Atlassian's $225M Bet
The Real Cost of AI Transformation. For CFOs and finance leaders: cost implications, budget planning, and ROI benchmarks from enterprise AI deployments.
March 12, 2026AI Models & PlatformsGPT-5.4 vs Claude Opus 4.6 for Coding: Which AI Writes Better Production Code?
Enterprise AI analysis: GPT-5.4 vs Claude Opus 4.6 for Coding. Strategic insights, ROI considerations, and implementation guidance for technical and business...
March 13, 2026Enterprise AIMicrosoft Bets $99/User on AI Agent Governance
Microsoft Bets $99/User on AI Agent Governance. For CFOs and finance leaders: cost implications, budget planning, and ROI benchmarks from enterprise AI deplo...
March 16, 2026AI AutomationClaude Can Now Loop: 5 Enterprise Workflows Now Automated
Claude's scheduled tasks feature lets enterprise teams automate recurring AI workflows without human oversight. Here's what platform teams are building first.
March 9, 2026AutomationGoogle Just Made AI Assistants Actually Useful for Work
Enterprise AI analysis: Google Just Made AI Assistants Actually Useful for Work. Strategic insights, ROI considerations, and implementation guidance for tech...
March 11, 2026Enterprise AICursor's Hidden Chinese Model: A Wake-Up Call for AI Vendor Due Diligence
Cursor launched Composer 2 without disclosing its Chinese model foundation. The $29.3B company's transparency failure exposes systemic gaps in enterprise AI procurement.
March 22, 2026AI InfrastructureGimlet Labs Raises $80M to Solve AI's Biggest Waste Problem
Stanford founder's multi-silicon cloud delivers 3-10x faster AI inference by orchestrating workloads across CPUs, GPUs, and specialized chips. Already running at top frontier labs and hyperscalers with 8-figure revenue.
March 23, 2026Enterprise AIOracle's Free Agentic Platform: No-Code AI Agents Built Into Your ERP
Oracle's free no-code agent builder includes ROI tracking, workflow orchestration, and governance—built into Fusion Applications.
March 24, 2026Enterprise AIOpenAI Turns ChatGPT Into a Sales Channel for Retailers
OpenAI launched Agentic Commerce Protocol expansion for product discovery — Target, Walmart, and Shopify are integrated. High-intent shoppers meet conversational AI.
March 25, 2026GranolaGranola's $125M Series C: Why Meeting Notes Just Became Enterprise Infrastructure
How Granola raised $125M at $1.5B valuation by pivoting from consumer app to enterprise AI platform — and what it means for CIOs choosing meeting intelligence tools.
March 25, 2026Enterprise AIGranola's $125M: Meeting Notes to Enterprise Platform
From $250M to $1.5B in under a year: How Granola turned invisible meeting transcription into an enterprise AI platform that enterprises actually deploy.
March 26, 2026Enterprise AIGoogle Agent Smith Writes 30% of Code While Humans Sleep
Google's Agent Smith now generates 30%+ of production code autonomously. Access got capped after demand surge. Context beats capability.
March 29, 2026Enterprise AIClaude Code Source Leaked: $2.5B Product Blueprint Exposed
Analysis of Claude Code Source Leaked. For enterprise leaders: strategic implications, cost considerations, and implementation guidance for AI decision-makers.
March 31, 2026Developer ToolsCodex Pay-As-You-Go: 6x Growth Proves Seat Fees Kill Adoption
OpenAI eliminates Codex seat fees after 6x usage surge since January. 2M+ weekly builders show pay-as-you-go beats fixed pricing for engineering tools.
April 3, 2026