Cursor

Cursor is an AI-powered software development platform built around coding agents that can understand repositories, modify code, execute commands, review changes, and complete increasingly complex engineering tasks. Originally developed by Anysphere as an AI-first code editor, Cursor has evolved into a broader agentic software-development platform and became part of SpaceX in August 2026.

The company serves individual developers, startups, engineering teams, and large enterprises through a combination of subscription plans, usage-based AI economics, proprietary coding models, third-party frontier models, cloud agents, and development-tool integrations.

Last updated: September 1, 2026

About Cursor

Cursor is a developer-software company focused on integrating artificial intelligence directly into the software-engineering workflow.

Rather than positioning AI as a separate chatbot that developers consult while programming, Cursor is designed around the idea that AI should operate inside the development environment itself. Its systems can work with repository context, inspect files, modify code, execute commands, review changes, and interact with tools developers already use.

This places Cursor within the emerging category of AI coding agents rather than conventional code-completion software alone.

The distinction is important. Earlier AI programming tools primarily predicted what a developer might type next or answered programming questions. Cursor increasingly allows developers to describe an objective—such as fixing a bug, refactoring part of a repository, or implementing a feature—and let an AI agent perform parts of the engineering work.

Cursor was created by Anysphere, which was founded in 2022 by Michael Truell, Sualeh Asif, Arvid Lunnemark, and Aman Sanger. Following rapid growth during the generative-AI boom, Anysphere was acquired by SpaceX in August 2026.

What Does Cursor Do?

Cursor develops AI-powered tools that help programmers understand, write, modify, debug, test, review, and manage software.

Its core activity is combining several layers that would traditionally operate separately:

• a software-development environment;

• repository and codebase context;

• AI coding models;

• autonomous or semi-autonomous agents;

• terminal and browser tools;

• code review;

• cloud development environments;

• and integrations with external development systems.

Repository-Aware Software Development

Cursor can work with information from an existing codebase rather than relying only on isolated code snippets.

This allows its agents to search repositories, identify relevant files, understand relationships between components, and make coordinated changes across several parts of a project.

That repository-aware architecture is one of the company's main distinctions from general-purpose AI chat tools used for programming.

AI Agents for Development Tasks

Cursor Agent can receive a broader software-development objective and use development tools to work toward it.

Depending on the task and configuration, an agent can inspect files, edit code, execute shell commands, use browser tools, evaluate errors, and continue iterating.

Cursor has expanded this model into Cloud Agents, where development tasks can be performed in isolated remote environments rather than entirely on the developer's local machine.

AI-Assisted Code Review

Cursor also operates in the software-review stage.

Bugbot applies AI to code review by inspecting changes for potential bugs, security issues, and code-quality problems.

This expands Cursor beyond code generation into software maintenance and engineering quality.

Cursor Business Model

Cursor operates primarily as a subscription software business with usage-based AI economics.

Individual developers pay for monthly plans that provide different levels of access to Cursor's AI features and model usage. Teams can purchase per-user plans with administration, shared controls, and organizational features, while larger companies can negotiate Enterprise arrangements.

Unlike traditional SaaS, however, Cursor's cost structure is directly affected by AI inference.

Every substantial model request consumes computing resources. Different AI models also have different inference costs, meaning the cost of serving two customers on the same subscription can vary considerably depending on how intensively they use agents and which models they select.

Subscription Revenue

Cursor offers free and paid individual plans together with Teams and Enterprise offerings.

Recurring subscriptions give the company predictable software revenue while allowing it to serve developers ranging from individual users to large organizations.

Usage-Based AI Economics

Subscription revenue is combined with model-usage allowances and additional consumption economics.

Heavy agent workloads, long development sessions, expensive third-party models, Cloud Agents, and automated review can consume significantly more AI resources than lightweight autocomplete.

Cursor therefore increasingly resembles a hybrid between conventional SaaS and AI infrastructure consumption.

Proprietary Models and Vertical Integration

Cursor also develops its own coding models.

Its proprietary model strategy is economically important because relying entirely on third-party AI providers would leave the company dependent on external model pricing, availability, and product decisions.

Developing coding-specific models gives Cursor greater control over latency, agent behavior, performance, and inference economics.

SpaceX ownership potentially strengthens this model by giving Cursor access to considerably larger computing infrastructure.

How Does Cursor Make Money?

Cursor generates revenue primarily from paid subscriptions and AI usage.

As of September 2026, its public plan structure includes free, individual, team, and enterprise offerings.

The currently listed individual pricing includes:

• Hobby: Free

• Start: ₹649 per month in India

• Pro: $20 per month

• Pro+: $60 per month

• Ultra: $200 per month

Cursor also lists Teams Standard at $40 per user per month and Teams Premium at $120 per user per month, while Enterprise pricing is customized.

Cursor's Revenue Growth

Before the SpaceX acquisition, Anysphere reported unusually rapid revenue growth.

In January 2025, the company said it had surpassed $100 million in annual recurring revenue.

By June 2025, Cursor reported more than $500 million in annual recurring revenue.

In November 2025, the company said it had crossed $1 billion in annualised revenue.

These figures were company-reported and reflected Cursor's growth before it became part of SpaceX.

A detailed public breakdown showing the percentage of Cursor revenue generated by individual subscriptions, team accounts, Enterprise contracts, or usage charges has not been provided in the material used for this profile. For that reason, exact revenue shares should not be inferred.

Products and Services

Cursor's product portfolio centers on its AI coding environment and the different agent, model, review, and integration layers built around it.

Cursor AI Coding Environment

Cursor's principal product is its AI-powered development environment.

It combines conventional software editing with repository-aware AI assistance, allowing developers to work with AI without continually moving code between an IDE and an external chatbot.

The environment serves as the interface through which Cursor's completion, Agent, model, review, and customization systems operate.

Cursor Agent

Cursor Agent performs higher-autonomy development work.

It can search repositories, inspect files, modify multiple files, execute commands, use development tools, and work through larger engineering objectives.

Agent represents one of Cursor's most important transitions from AI-assisted editing toward AI-executed software-development tasks.

Cloud Agents

Cloud Agents extend Cursor's agent system into remote development environments.

These agents work inside isolated virtual machines containing cloned repositories and the development resources required to complete assigned tasks.

Cloud Agents allow software work to continue without requiring the developer's local computer to remain actively involved throughout the task.

Multi-Agent Development

Cursor supports multiple agents working in parallel.

Parallel-agent workflows allow independent development tasks to be delegated simultaneously, using isolated environments designed to prevent changes from simply overwriting one another.

This model is particularly relevant to Cursor's broader vision of developers supervising increasing amounts of AI-generated implementation work.

Cursor Tab

Cursor Tab provides AI-powered code completion.

It represents the lower-autonomy side of Cursor's product portfolio, predicting code and edits while developers remain directly involved in the implementation process.

Tab remains useful for fast, incremental coding tasks where a full autonomous agent would be unnecessary.

Composer and Cursor Models

Cursor develops proprietary models for software-engineering workloads.

Its current first-party model portfolio includes Composer and Cursor Grok variants.

These models allow Cursor to optimize AI behavior specifically for agentic coding, tool usage, file editing, and development workflows rather than relying exclusively on general-purpose models.

Third-Party AI Models

Cursor is also a multi-model platform.

As of September 2026, its documentation lists models from providers including OpenAI, Anthropic, Google, and SpaceXAI alongside Cursor's own models.

This gives developers access to different model families for different workloads while reducing reliance on a single external provider.

Cursor Router

Cursor Router is designed to select models automatically according to factors such as complexity, intelligence requirements, and cost.

Its routing strategies include Cost, Balance, and Intelligence.

As of September 1, 2026, Cursor Router is available for Teams and Enterprise accounts, with broader availability described as part of Cursor's future rollout.

Bugbot

Bugbot is Cursor's AI-assisted code-review system.

It reviews code changes for potential bugs, security concerns, and quality issues and can participate in pull-request workflows.

Bugbot is usage-based rather than simply an unlimited feature bundled into every paid subscription.

Browser, MCP, Rules, Skills, and Plugins

Cursor also provides infrastructure for extending and customizing agent behavior.

Browser tools can help agents inspect and test web applications. Model Context Protocol integrations allow agents to connect with external tools and data. Rules and skills let developers or organizations provide persistent instructions and repeatable workflows.

Plugins can combine several of these capabilities into reusable development configurations.

These systems are important to Cursor's enterprise strategy because companies often need AI-generated work to follow internal coding standards, architecture rules, and development processes.

Customers and Markets

Cursor serves several overlapping customer groups within the software-development market.

Individual Developers

Individual programmers are a major part of Cursor's user base.

The product is used for everyday programming tasks including repository navigation, code completion, debugging, testing, refactoring, and AI-assisted implementation.

Cursor's free and paid individual plans are structured around different levels of AI usage.

Startups and Software Teams

Startups and engineering teams use Cursor where developers need to move quickly across features, bugs, tests, documentation, and repository maintenance.

The platform's agent workflows are particularly relevant to teams experimenting with delegating repeatable or clearly scoped engineering work to AI.

Enterprise Customers

Cursor also targets large organizations through Teams and Enterprise products.

Organizational capabilities include areas such as centralized administration, billing, privacy controls, analytics, shared resources, authentication, code review, and AI governance.

By June 2025, Cursor said the platform was used by more than half of the Fortune 500, including companies such as NVIDIA, Uber, and Adobe.

That figure is company-reported, but it illustrates how quickly Cursor expanded from an individual developer tool into enterprise software.

Geographic Market

Cursor is fundamentally a global software product because its customers are developers and engineering organizations rather than location-dependent consumers.

The material reviewed for this profile does not provide a reliable geographic revenue breakdown, so no regional revenue shares should be inferred.

Competitors

Cursor competes across several layers of the AI software-development market rather than against only one product category.

AI Coding Assistants

GitHub Copilot is one of Cursor's clearest major competitors.

Both products aim to integrate generative AI directly into development workflows, although Cursor has increasingly differentiated itself through repository-aware agents, Cloud Agents, proprietary models, and broader agent orchestration.

Agentic Coding Platforms

Cursor also competes with other platforms designed around autonomous or semi-autonomous software-development agents.

Competition in this category increasingly depends on how effectively an AI system can understand repositories, use tools, execute code, verify its work, and operate with limited supervision.

General-Purpose AI Assistants

General-purpose frontier AI systems also compete indirectly with Cursor because developers frequently use them for code generation, debugging, architecture questions, and software explanations.

Cursor's strategic response is deeper integration with the development environment rather than competing only on conversational model quality.

Model-Level Competition

Cursor also competes at the AI-model layer.

By developing its own models while continuing to support third-party providers, Cursor is attempting to avoid dependence on a single model company and gain greater control over performance and costs.

Company History

Cursor's history is closely tied to the rapid shift from AI code completion toward agentic software development.

The most important historical transition was not any single funding round. It was Cursor's shift from an AI-first editor into an environment where developers could increasingly delegate complete software-development tasks to agents.

Leadership

Cursor was founded by Michael Truell, Sualeh Asif, Arvid Lunnemark, and Aman Sanger, who met while studying at MIT.

Michael Truell became Cursor's CEO and its most visible public executive as the company expanded from an early-stage AI editor into a major developer-AI platform.

The founding team's background in mathematics, computer science, and artificial intelligence helped shape Anysphere's research-driven approach to software development.

Following the SpaceX acquisition, Cursor now operates inside a much larger technology and AI infrastructure organization. The available material does not provide a complete post-acquisition executive structure, so it would be inappropriate to invent additional leadership roles or reporting relationships.

Ownership, Subsidiaries and Corporate Structure

Cursor was originally developed and operated by Anysphere, a privately funded technology company.

That structure changed in 2026.

On August 14, 2026, Cursor announced that Anysphere had officially become part of SpaceX following an earlier SpaceXAI partnership centered on AI model training.

Cursor therefore no longer operates as an independent venture-backed company in the way it did during its 2024 and 2025 fundraising rounds.

The Cursor product continues to operate as a developer-facing software platform within the wider SpaceX ecosystem.

Before the acquisition, Anysphere had raised several major financing rounds:

• $60 million Series A in 2024;

• $105 million Series B in January 2025;

• approximately $900 million Series C in June 2025;

• and $2.3 billion Series D in November 2025.

The Series D valued the business at $29.3 billion post-money.

No significant Cursor subsidiaries or minority holdings are identified in the research used for this profile, so additional corporate entities should not be assumed.

Cursor Strategy

Cursor's current strategy centers on moving software development from AI-assisted typing toward AI-executed engineering work.

Several priorities support that direction.

Expanding From Editor to Coding-Agent Platform

Cursor's most important product strategy is increasing the amount of software-development work that AI can perform directly.

Code completion remains part of the product, but the company increasingly emphasizes agents that can understand tasks, inspect repositories, modify software, use development tools, and verify results.

This changes Cursor's competitive position from an intelligent editor toward a broader software-engineering platform.

Building Proprietary Coding Models

Cursor's investment in Composer and other proprietary models is both a technology strategy and an economic strategy.

Coding-specific models can potentially be optimized around the particular tool calls, editing behavior, latency requirements, and inference economics of Cursor's platform.

This gives the company more control than it would have if every AI interaction depended on outside providers.

Maintaining a Multi-Model Platform

Cursor continues to support third-party frontier models alongside its own.

The multi-model approach lets the company offer different models for different workloads and reduces dependence on the capabilities of one external provider.

Model routing is an extension of this strategy, allowing the platform to balance cost, speed, and reasoning ability at the orchestration layer.

Expanding Remote and Parallel Agent Work

Cloud Agents and parallel agents point toward a future where developers supervise more work than they manually implement.

Instead of one developer interacting with one AI suggestion at a time, Cursor is building toward environments in which multiple software tasks can be delegated to independent agents and reviewed afterward.

Deepening Enterprise Adoption

Enterprise adoption is another important growth priority.

As AI agents receive greater access to repositories, terminals, secrets, and company systems, larger customers need governance, privacy controls, administration, authentication, auditability, and security management.

Cursor's Teams, Enterprise, Bugbot, rules, skills, and integration layers support this broader organizational use case.

Using SpaceX Infrastructure for Greater Vertical Integration

SpaceX ownership adds another strategic layer.

Cursor has said that joining SpaceX provides access to large-scale computing infrastructure that can support development of stronger and more economically efficient AI models.

If successful, that could reduce Cursor's dependence on external infrastructure while improving its ability to optimize models specifically for software engineering.

Key Challenges and Risks

Cursor's growth does not remove several important technical, economic, security, and competitive risks.

AI-Generated Code Can Be Wrong

Coding agents cannot guarantee that generated software is correct.

Models can misunderstand requirements, introduce subtle bugs, choose outdated approaches, or generate insecure implementations.

The risk becomes more significant as agents make larger changes across more files because a developer must review a greater amount of AI-generated work.

Cursor therefore remains an engineering accelerator rather than a substitute for software-engineering judgment.

Large Repositories Create Context Problems

Repository awareness is a major Cursor strength, but large codebases can contain hidden relationships, undocumented assumptions, legacy behavior, and architectural dependencies that are difficult for an AI system to identify perfectly.

An agent working from incomplete context can make changes that appear correct locally while creating problems elsewhere in the system.

Autonomous Agents Expand Security Risk

Security becomes more complex when AI can execute commands and interact with repositories rather than merely suggest text.

Cloud Agents can operate remotely and require access to development resources, while local agents may also interact with terminals, browsers, files, and external services.

This creates risks involving:

• credentials;

• source-code confidentiality;

• network access;

• prompt injection;

• destructive commands;

• insecure generated code;

• and excessive permissions.

Cursor provides privacy, security, network, and administrative controls, but organizations still need their own access policies and review processes.

Heavy Agent Usage Is Expensive

AI coding becomes economically different from conventional software subscriptions when customers use long-running agents.

Autocomplete is relatively lightweight.

Agents that repeatedly reason, search repositories, edit files, execute tools, and use high-cost models can consume substantially more inference resources.

Cursor must therefore balance user expectations for generous AI usage against the underlying cost of operating the service.

Dependence on External Model Providers

Cursor's multi-model architecture creates flexibility but also dependence.

Third-party AI providers can change pricing, access terms, model availability, or strategic relationships.

This risk became particularly visible in August 2026, when Reuters reported that OpenAI planned to stop supplying models to Cursor on November 12, 2026, although discussions between the companies were reportedly continuing.

As of September 1, 2026, OpenAI models remained listed in Cursor's documentation.

Cursor's proprietary model strategy reduces this dependency but does not eliminate it.

Competition Is Moving Quickly

AI coding is one of the most competitive areas of generative AI.

Cursor competes not only with established coding assistants such as GitHub Copilot but also with emerging agent platforms, model providers, and development tools adding autonomous capabilities.

Features that currently differentiate Cursor can become standard across competing products, making model quality, workflow design, infrastructure, security, and economics increasingly important.

Key Facts About Cursor

• Cursor was developed by Anysphere, which was founded in 2022 by Michael Truell, Sualeh Asif, Arvid Lunnemark, and Aman Sanger.

• Cursor became part of SpaceX on August 14, 2026, following an earlier SpaceXAI model-training partnership.

• Cursor has evolved from an AI-first editor into a broader coding-agent platform capable of working across repositories and development tools.

• Cursor operates a multi-model architecture, combining proprietary Cursor models with models from external frontier AI providers.

• Cloud Agents allow AI development tasks to run inside remote isolated environments rather than only on a developer's local computer.

• Cursor supports parallel agent workflows, allowing several software tasks to be delegated simultaneously when the work can be separated effectively.

• Cursor reported that it was used by more than half of Fortune 500 companies by June 2025.

• Cursor currently states that it holds SOC 2 Type II, ISO/IEC 27001:2022, ISO/IEC 42001:2023, and AIUC-1 security and AI-governance certifications.

• Cursor's strategic direction increasingly treats AI as a software-engineering participant rather than only a suggestion engine.

Frequently Asked Questions

What Is Cursor AI?

Cursor AI is an AI-powered software-development platform that combines code editing, repository context, AI models, and coding agents. Its tools can help developers understand codebases, modify files, execute commands, review changes, and complete broader engineering tasks.

Who Owns Cursor?

Cursor is part of SpaceX. Cursor announced on August 14, 2026 that Anysphere, the company that developed Cursor, had officially become part of SpaceX.

Who Founded Cursor?

Cursor was developed by Anysphere, founded in 2022 by Michael Truell, Sualeh Asif, Arvid Lunnemark, and Aman Sanger.

How Does Cursor Make Money?

Cursor primarily makes money from software subscriptions and AI usage. Revenue comes from individual plans, team subscriptions, Enterprise agreements, model consumption, and products such as usage-based Bugbot code review.

Is Cursor AI Free?

Yes. Cursor currently offers a free Hobby plan with limited access to features including Agent, Chat, and Tab. Paid individual plans start with Pro at $20 per month as of September 1, 2026.

What AI Models Does Cursor Use?

Cursor supports its own models as well as models from external providers. As of September 2026, its documentation lists providers including OpenAI, Anthropic, Google, and SpaceXAI alongside Cursor models such as Composer and Cursor Grok variants.

Does Cursor Still Use OpenAI Models?

Yes, as of September 1, 2026, Cursor's documentation still lists OpenAI models.

Reuters reported on August 29, 2026 that OpenAI planned to stop providing models to Cursor on November 12, 2026, although discussions between the companies were continuing. The situation is therefore time-sensitive.

What Is the Difference Between Cursor Tab and Cursor Agent?

Cursor Tab is primarily an AI code-completion system that predicts edits while a developer works.

Cursor Agent can take a broader objective, search a repository, modify files, execute commands, and perform more of the development workflow independently.

What Are Cursor Cloud Agents?

Cloud Agents are remote Cursor agents that operate inside isolated cloud development environments. They can work on software tasks without requiring the developer's local computer to remain actively involved throughout the process.

Is Cursor Safe for Company Code?

Cursor provides features such as Privacy Mode, administrative controls, security certifications, isolated Cloud Agent environments, network policies, and other security mechanisms.

However, organizations should still evaluate repository permissions, secrets, external model providers, network access, generated changes, and internal AI governance before giving autonomous agents access to sensitive systems.

Sources:

• Cursor Agent Documentation

• Cursor Pricing Documentation

• Cursor Models and Pricing Documentation

• Cursor Cloud Agent Documentation

• Cursor Cloud Agent Security Documentation

• Cursor Cloud Agent Network Security Documentation

• Cursor Privacy Documentation

• Cursor Security and Compliance Page