Cursor Review 2026: Plans, Models and Who It Suits After the SpaceX Deal
Disclosure: Cursor does not pay us, and the Cursor links on this page are not affiliate links. The three Amazon links near the end are. Our affiliate disclosure explains how the site is funded.
The short version: Cursor is still the best-known AI-first code editor, but it is a different product from the one most reviews describe. In August 2026 it became part of SpaceX. Its plans were rebuilt around a $8 Plus tier that runs only Cursor’s own Grok and Composer models, plus higher tiers from $30 that add Claude, GPT and Gemini at API list prices. If you want Cursor’s editor with the models it was known for, budget for Premium at $30 a month, not the old $20. If you mainly want autocomplete in the editor you already use, GitHub Copilot at $10 is the cheaper route.
How we reviewed it
We did not run Cursor on our own projects, and we do not publish scores of our own. Everything below comes from Cursor’s own pricing page, documentation, security page and blog, read on October 27, 2026. Where something is a judgment rather than a published fact, we say so.
What is Cursor?
Cursor is a standalone code editor. Its documentation says it “is based upon the VS Code codebase” (Cursor docs), so VS Code themes, keybindings and settings carry over. Extensions install from the Open VSX registry rather than Microsoft’s marketplace. It runs on Mac, Windows and Linux.
On August 14, 2026, Cursor’s blog announced that “Cursor has officially been acquired by SpaceX”, completing a process that began in April (Cursor blog). The effects show on the pricing page, which now bundles X perks into the plans and puts Grok models at the bottom tier.
Key features
Agent
Cursor’s Agent is, in its own words, an assistant that can “complete complex coding tasks independently, run terminal commands, and edit code” (Agent docs). You describe a change, it edits across files and shows you diffs to accept or reject. Older reviews call this “Composer”. That name now belongs to Cursor’s own coding model, Composer 2.5, which Cursor describes as its most price-efficient coding model (Composer).
Tab
Tab is Cursor’s autocomplete. It predicts your next edit from your recent changes, the surrounding code and linter errors, not just the next few characters (Tab help).
Cloud agents
Cloud agents, formerly called Background Agents, run tasks in isolated cloud machines. You can launch them from the desktop app, the web, iOS, Slack, GitHub, Bitbucket, Linear or the API. They need the Premium plan or higher (cloud agent docs).
Bugbot
Bugbot reviews pull requests for bugs, security issues and code-quality problems on GitHub, GitLab, Bitbucket and Azure DevOps. It is billed by usage and comes with Premium and up (Bugbot docs).
CLI and JetBrains
A terminal agent, started with the agent command, runs on macOS, Linux, WSL and Windows PowerShell (CLI docs). JetBrains users are no longer shut out entirely: Cursor’s agent can run inside JetBrains AI Chat through the Agent Client Protocol. That requires a paid Cursor plan and JetBrains AI Assistant 2025.1 or later (JetBrains integration). It is not a full Cursor plugin, and Tab does not come with it.
Rules and code search
You can steer the agent with project rules (.cursor/rules), user rules, team rules and an AGENTS.md file (rules docs). For search, Cursor’s docs describe a local index, and say Cursor “does not store embeddings of your codebase for search” (search docs).
Cursor pricing (October 2026)
| Plan | Price | Models | Usage |
|---|---|---|---|
| Free | $0 | Cursor’s own models; no third-party models | “Limited agent requests”, “Limited Chat requests” |
| Plus | $8/mo | “Only Grok models” | Base weekly pool; includes X perks |
| Premium | $30/mo | Adds Claude, GPT, Gemini and others at API list price | 4x Plus; cloud agents and Bugbot |
| Super | $100/mo | Same as Premium | 15x Plus |
| Ultra | $200/mo | Same as Premium | 50x Plus |
| Teams | From $30/user/mo | Team controls | Privacy Mode on by default |
Sources: Cursor pricing, pricing docs and models and pricing, read October 27, 2026. The pricing page showed monthly prices only that day.
The model line-up on paid tiers is wide. Cursor’s own Grok 4.5 to 4.7 models and Composer 2.5 sit alongside Anthropic’s Claude models up to Opus 5.5 and Fable 5.1, OpenAI’s GPT-5.x family, Google’s Gemini 3.x and Gemini 4 Argon, and models from DeepSeek, GLM, Kimi and Meta. The catch is that usage is one weekly pool shared across Cursor and Grok products, with no published request count, and third-party models are charged at their API list price on top.
Privacy
Cursor holds a SOC 2 Type II attestation and offers Privacy Mode on every plan, including Free. With it on, Cursor says “we will not train on your data” (Cursor security). With it off, its data-use page says Cursor may use and store codebase data and prompts to improve its features and train its models (data use). Privacy Mode is on by default for Teams and Enterprise. Cursor does not state a default for individual plans, so check the setting before you open client code.
What we like
- An editor built around an agent, with VS Code settings and themes carrying over
- The widest model menu of any editor on the paid tiers
- Cloud agents and PR review (Bugbot) from Premium up
- Privacy Mode available on every plan, with a SOC 2 Type II attestation
What could be better
- The $8 Plus plan runs only Cursor’s own models
- No published request counts; usage is a weekly pool
- Third-party models billed at API prices on top of the plan
- Extensions come from Open VSX, not Microsoft’s marketplace
Who should use Cursor?
A good fit: developers who want the editor itself built around an agent, are happy to leave VS Code for a close relative of it, and will pay for Premium to get Claude, GPT and Gemini. Teams that want cloud agents and automated PR review in the same product also fit.
Look elsewhere if: you mainly want autocomplete in your current IDE (Copilot covers VS Code, Visual Studio, JetBrains, Vim and Neovim, Xcode and Eclipse); you want a coding agent bundled with a $20 chat plan (Claude Code or Codex, compared in our Claude vs ChatGPT for coding piece); or you need a published usage allowance before you buy.
Cursor vs GitHub Copilot
| Feature | Cursor | GitHub Copilot |
|---|---|---|
| Entry paid price | $8 (Grok only); $30 for third-party models | $10 (Pro, 1,500 AI Credits) |
| Autocomplete | Tab, inside Cursor | Unlimited on paid plans, in most major IDEs |
| Editors | Cursor’s own editor; agent in JetBrains via ACP | VS Code, Visual Studio, JetBrains, Vim/Neovim, Xcode, Eclipse |
| Agent that opens PRs | Cloud agents (Premium+) | Copilot cloud agent (from Pro) |
| PR review | Bugbot (Premium+, usage-billed) | Copilot code review (from Pro) |
| Training on your code | Not with Privacy Mode on | Free/Pro/Pro+ interaction data used unless you opt out |
For the full field, including Claude Code, Codex, Devin Desktop and Tabnine, see our best AI coding assistants comparison.
Frequently asked questions
Is Cursor free?
There is a free plan with “limited agent requests” and “limited Chat requests” on Cursor’s own models. Third-party models such as Claude and GPT start on Premium at $30 a month.
Who owns Cursor?
SpaceX. Cursor’s blog said on August 14, 2026 that it “has officially been acquired by SpaceX”.
Can I use Claude in Cursor?
Yes, on Premium and above, billed at Anthropic’s API list price within Cursor’s usage pool. The $8 Plus plan runs only Cursor’s own models.
Does Cursor work in JetBrains?
Partly. Cursor’s agent can run inside JetBrains AI Chat through the Agent Client Protocol on a paid Cursor plan. The full editor, including Tab, is Cursor’s own app.
Does Cursor train on my code?
Not with Privacy Mode on, which is available on every plan. With it off, Cursor’s data-use page says it may use codebase data and prompts to train its models.
Our verdict
Cursor is still the most complete AI-first editor, with an agent, autocomplete, cloud agents and PR review in one product. But the old “$20 and you get everything” pitch is gone. The cheap tier is a Grok-only plan, and the tier most developers will want, with Claude, GPT and Gemini, starts at $30 plus API-priced usage. Try the free plan to see whether the editor suits you, read the model line on the pricing page before you upgrade, and compare it against Copilot Pro at $10 and the $20 chat plans that now include a coding agent.
About the author
Lennit Herrera is the founder and editor of AI Tool Stack Reviews, and he builds and audits websites for a living. We do not run hands-on tests and we do not publish scores of our own. Prices, limits and features come from the vendor’s own pages and are dated. Corrections are welcome through our contact page.
Gear we’d pair with an AI editor
A faster AI editor is most useful when the rest of your setup keeps up. A few things worth having:
- A good mechanical keyboard, because you’ll be accepting and editing a lot of diffs.
- An ultrawide monitor, to keep diffs and the agent panel side by side.
- The Pragmatic Programmer. AI writes the code faster, but you still own the judgment.
As an Amazon Associate we earn from qualifying purchases.
