Anysphere launched Cursor Ultra on June 16, 2025—not in August 2026. The $200-a-month plan was designed for developers who use AI coding agents heavily. Cursor initially marketed it as offering 20 times the usage of its Pro plan, but that figure was never a guarantee of unlimited coding or a fixed number of requests.
Under Cursor’s current documented pricing model, Ultra includes $400 of API agent usage plus bonus usage, with the amount of work that buys varying by model, token consumption, context size, and features such as Max Mode. For most occasional users, Pro or Pro+ is likely the more sensible financial choice.
As an Amazon Associate I earn from qualifying purchases.
What Anysphere announced
Anysphere, the company behind the Cursor AI coding editor, announced Cursor Ultra on June 16, 2025. TechCrunch reported the launch on June 17, 2025. The headline price was $200 per month.
Cursor positioned Ultra as a power-user subscription with:
#1 Best Overall
- 20 times more usage than Cursor Pro, according to the launch announcement;
- priority access to new features; and
- access to models from providers including OpenAI, Anthropic, Google and xAI, as described by Cursor.
The purpose was to give heavy users a more predictable subscription than separately managing multiple model subscriptions or unpredictable API bills. “Predictable” did not mean unlimited: the plan still depended on usage limits, model pricing and account settings.
Cursor also said at launch that Pro would move toward an unlimited-with-rate-limits model and that tool-call limits would be removed by default. Existing users could retain the earlier 500-request-limit approach through a dashboard setting. Those launch-era details should not be confused with Cursor’s later usage-based pricing documentation.
Read the original Cursor announcement and TechCrunch’s launch report.
Quick wins for a faster PC:
Clear out junk files and repair common Windows errorsFree Scan →Scan for outdated or missing drivers - takes under a minuteDriver Scan →Repair Windows errors before they cause bigger problemsFix Now →What Cursor Ultra includes under the current pricing model
Cursor’s current documentation describes Ultra as a usage-based subscription rather than a limitless pass. The documented individual-plan figures are:
| Plan | Monthly price | Included API agent usage |
|---|---|---|
| Pro | $20 | $20 plus bonus usage |
| Pro+ | $60 | $70 plus bonus usage |
| Ultra | $200 | $400 plus bonus usage |
These are U.S.-dollar headline prices. Taxes, regional billing, plan names, allowances and availability can change, so check the official Cursor pricing page and pricing documentation before subscribing.
Ultra also includes unlimited tab completions, extended agent usage limits on supported models, and access to Bugbot. Bugbot has separate product pricing, so inclusion in the broader Cursor ecosystem should not be interpreted as every Bugbot capability being free under the $200 subscription.
Background Agents are another important exception. Cursor says they are charged at API pricing and require the user to set a spending limit. An Ultra subscription therefore does not automatically cover unlimited long-running background work.
What the $400 allowance means
The $400 figure is an included API agent usage allowance valued according to inference pricing. It is not cash, a refundable balance, or a universal number of requests.
Cursor’s documentation gives approximate median-equivalent examples for Ultra of:
- about 4,500 Sonnet 4 requests;
- about 11,000 Gemini requests; or
- about 10,000 GPT-5 requests.
These figures are estimates, not quotas. Actual consumption depends on the model, input and output tokens, cached input, repository context, reasoning length, and the complexity of the task. A short code edit and a large multi-file refactor do not consume the same amount of capacity.
Max Mode can provide larger context windows—up to 1 million tokens for certain models—but it can also consume usage faster. It may be useful for a large repository while being unnecessarily expensive for routine edits.
The Tool Desk
Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Why “20× more usage” is easy to misunderstand
“20×” was launch-era positioning relative to Pro. It does not mean:
- 20 times faster development;
- 20 times better output;
- 20 times as many requests regardless of model; or
- unlimited access to every model and feature.
The practical value depends on what a developer does. A user who mostly accepts tab completions may never approach the allowance. A developer running several agents against a large codebase can consume it much faster, particularly when using expensive reasoning models, long prompts or Max Mode.
Cursor’s dashboard can show usage and token breakdowns. That information is more useful than counting requests because it reveals which models and workflows are driving the bill.
Can Ultra create extra charges?
Yes. Included usage is predictable, but total spending may not be. After included usage is exhausted, Cursor says users can purchase additional usage at cost or upgrade where applicable. Background Agents are separately charged at API pricing and have their own spending-limit implications.
Do these 3 things before closing this tab:
1Fix the driver behind crashes, sound loss and screen glitches2Repair Windows errors before they cause bigger problems3Scan for outdated or missing drivers - takes under a minuteBefore enabling overages, check:
- whether additional usage is enabled;
- what spending limit applies;
- whether Background Agents are running;
- which models are being used; and
- whether a team or employer account has separate billing controls.
This distinction matters for budgeting: a $200 subscription can become a larger monthly technology expense if overages are allowed without a cap.
Who might get value from Ultra?
Ultra is most defensible when Cursor is a primary work environment and the user can demonstrate sustained, high-volume agent usage. Possible candidates include:
- developers working with agents throughout the day;
- users running multiple parallel agents or automation jobs;
- engineers who regularly exceed Pro+ allowances;
- people who need to switch among models from several providers inside one editor; and
- individuals or businesses already spending around $200 a month across several coding tools, subscriptions or API accounts.
Cursor’s own usage guidance says daily tab users generally remain within about $20 of monthly usage, limited agent users often remain within the Pro allowance, daily agent users may reach roughly $60–$100 in monthly usage, and power users running multiple agents or automation can reach $200 or more. These are Cursor’s estimates, not an independent spending benchmark.
Rank #4
Who should probably avoid it?
Ultra is likely poor value for someone who:
- uses AI coding assistance only occasionally;
- primarily wants autocomplete;
- rarely uses agents;
- does not regularly consume Pro+ capacity;
- wants one provider’s model through a standalone chat or terminal interface; or
- needs a hard ceiling on total monthly spending.
It is also not the same as Cursor Teams or Enterprise. Organizations needing centralized procurement, governance, audit controls or compliance features should evaluate the relevant business plans rather than assuming an individual Ultra account meets those requirements.
Free tools Windows power users keep installed
One-click scans. No signup required.
A practical break-even test
Before upgrading, review the last one or two months of actual usage and answer these questions:
- Did you reach your plan’s included limit?
- Did you pay for additional usage?
- Which models consumed the most capacity?
- Did you use Background Agents or Max Mode?
- Are you paying separately for Claude, ChatGPT, Copilot or model APIs?
- Would Cursor’s editor integration still be worth $200 if its model mix or limits changed?
If you are not regularly approaching Pro+ usage, the data does not support moving straight to Ultra. Start with a lower tier, monitor consumption and upgrade only when the additional capacity solves a real constraint.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Cursor Ultra versus Claude Max
Anthropic lists Claude Max 5× at $100 per month and Max 20× at $200 per month. The plans include Claude Code and higher usage than Claude Pro, subject to Anthropic’s limits. See Anthropic’s pricing page and plan guide.
Claude Max may be a better fit for a terminal-first developer or someone committed to Anthropic’s models. Cursor is more attractive to users who want an editor-native workflow and the ability to choose among providers. The two $200 plans do not provide equivalent quotas or interfaces, so the price alone is not a meaningful comparison.
Recommended Free Tools
Anthropic also offers additional usage bundles for some plans. Those bundles are extra purchases and are subject to their own limits and pricing changes.
Best Value
Cursor Ultra versus ChatGPT Pro and Codex
OpenAI lists ChatGPT Pro at $200 per month and provides Codex access on applicable plans. OpenAI’s current Codex documentation describes token-based credits or rate-card accounting, making a direct comparison with Cursor’s $400 API agent allowance unreliable.
ChatGPT Pro may make more sense for developers already centered on OpenAI’s ecosystem or who want ChatGPT features alongside Codex. Cursor is better suited to users who specifically want an AI-first code editor with multi-provider model access. Consult OpenAI’s Pro help page and Codex rate card for current terms.
Where GitHub Copilot fits
GitHub Copilot offers individual and organizational tiers, including Free, Pro, Pro+, Max, Business and Enterprise options. Its strengths are GitHub repository integration, pull requests, code review and enterprise administration. GitHub’s current plans page also identifies access to third-party agents such as Claude Code and Codex.
Copilot may be the better financial and operational choice for a GitHub-centered team. Cursor remains more directly focused on the AI-native editor experience. Check GitHub’s live plan page for current prices and AI-credit allowances before comparing costs.
Important caveats before subscribing
- The date matters: Ultra launched on June 16, 2025. It is not a new August 2026 launch.
- Terms can change: AI plans frequently change model access, rate limits, credits and pricing.
- Unlimited needs a label: Tab completions may be unlimited while agent usage remains governed by allowances and model consumption.
- Bonus usage is not a guaranteed unlimited quota: Treat it as additional capacity under Cursor’s current terms.
- Annual pricing should be checked directly: Do not assume an annual discount or fixed annual Ultra price from older community discussions.
- Taxes and geography matter: The $200 figure is a headline U.S.-dollar price and may not equal the final billed amount in every region.
The Bottom Line
Bottom line: Cursor Ultra can make financial sense for a developer who uses Cursor agents heavily, needs multiple models in one editor and regularly approaches lower-tier limits. For occasional users, tab-completion users and anyone seeking a guaranteed hard spending cap, Pro or Pro+ is the safer starting point. Compare your actual usage—not the launch-era “20×” slogan—before committing to $200 a month.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
