Cursor Plans and Usage Limits 2026: What Each Tier Actually Gets You
· 9 min read

Cursor Plans and Usage Limits 2026: What Each Tier Actually Gets You

Cursor's current plans explained in plain English: what each tier includes, how the shared weekly pool works, and why absolute usage counts stay unpublished.

Cursor Plans and Usage Limits 2026: What Each Tier Actually Gets You

Last verified: 2026-11-15

PlanPricePublished limitsSource
Free$0/mo”Limited” agent and chat requests; absolute counts not published; no third-party modelsPricing, Docs
Plus$8/mo; $80/yrBase weekly pool shared across Cursor and Grok; absolute counts not publishedPricing, Docs
Premium$30/mo; $300/yr4x PlusPricing, Docs
Super$100/mo; $1,000/yr15x PlusPricing, Docs
Ultra$200/mo; $2,000/yr50x PlusPricing, Docs
Teams$30, $100, or $200 per seat/moMatches the seat’s tier (grid: “4–50x Plus”)Pricing, Docs
EnterpriseCustom”Pooled”Pricing, Docs

Prices exclude tax.

Direct answer: Cursor’s self-serve plans are Free at $0, Plus at $8/month, Premium at $30/month, Super at $100/month, and Ultra at $200/month. Paid plans get 4x Plus usage (Premium), 15x (Super), or 50x (Ultra) from one weekly-refreshing pool. Absolute request counts are not published.

Cursor is sold as part of The X Subscription, one subscription covering Grok, Cursor, and X. This article covers Cursor usage only. Every figure below comes from the pricing page and the docs pricing page, both fetched 2026-11-15.

Free: what $0 gets you

Direct answer: Free costs $0 per month and includes limited agent and chat requests. The exact counts are not published. Third-party models are not available.

The docs table marks Free usage as “Limited” and third-party models as “Not available” (Docs). The pricing page lists “Get started with” Limited AI usage, Limited agent requests, and Limited Chat requests (Pricing).

Plus: the $8 entry point

Direct answer: Plus costs $8 per month, or $80 per year. It gives a base weekly pool shared across Cursor and Grok, using only Cursor’s own models: Grok 4.7, Grok 4.6, Grok 4.5, and Composer 2.5. Third-party models are not included.

The models run at fixed settings: Fast is off, effort is medium, and Grok 4.7 uses the 256k context window (Docs). Plus excludes Auto, Cloud Agents, Automations, Bugbot, Origin code hosting, third-party models, and bring-your-own API keys. It includes Grok Chat, Voice, Imagine, Build, and Grok API access.

One wording note: the pricing grid calls the Plus pool “Basic” (Pricing), while the docs call it “Base” (Docs). No number attaches to it on either page.

Premium: the first real step up

Direct answer: Premium costs $30 per month, or $300 per year, and includes 4x the usage of Plus. That is a multiplier, not an absolute number. It adds Cloud Agents and Automations, code hosting on Origin, Bugbot code reviews, third-party models at API list price, and Grok Bot.

The pricing page labels Premium “Most popular” (Pricing). Premium is the cheapest tier where Claude, GPT, and Gemini models become usable. They are billed separately at each model’s API list price, paid from prepaid credit rather than from the weekly pool.

Super: priority at peak times

Direct answer: Super costs $100 per month, or $1,000 per year, and includes 15x Plus usage. On top of Premium, it adds “priority model access at peak times.” The absolute request count is not published.

That phrase is the only addition over Premium on the pricing page (Pricing). The docs describe the same tier as “priority model access at peak times” with no further detail (Docs).

Ultra: the top self-serve tier

Direct answer: Ultra costs $200 per month, or $2,000 per year, with 50x Plus usage, the largest published multiplier. Everything in Super carries over, plus earliest access to all products and models. The absolute usage behind the 50x figure is unpublished.

The docs describe Ultra as suiting power users “running several agents or Automations” (Docs).

Teams and Enterprise

Direct answer: Teams seats cost $30, $100, or $200 per seat per month, matching the Premium, Super, and Ultra tiers. You can mix seat tiers on one team, but Plus is not offered for teams. Enterprise is custom-priced with pooled usage across Grok and Cursor.

Each member gets the same weekly usage as the matching individual plan (Docs). Teams adds centralized billing, usage analytics, a team marketplace, and SAML/OIDC single sign-on. Enterprise adds invoicing, SCIM account provisioning, granular spend controls, and priority support.

On Teams and Enterprise, eligible third-party requests carry a Cursor Token Rate of $0.25 per million tokens on top of model pricing. The rate does not apply to Cursor models, to Cost mode, or to third-party models whose cache-read price is strictly below $0.10 per million tokens (Docs). Individual plans do not pay it.

How the shared weekly pool works

Direct answer: Every paid plan has one weekly pool of usage for Cursor models, shared across the editor Agent, the CLI, Cloud Agents, and Grok Chat, Imagine, Voice, and Grok Bot. It refreshes weekly, so one heavy day does not use up your month.

Different models draw from the pool at different rates, but neither page publishes per-product draw rates, so you cannot calculate your request count in advance. Remaining usage and the next reset date appear in your editor settings and under Usage at grok.com/dashboard (Docs). The Grok API is billed separately and does not draw from the pool.

Top-ups, Auto Top-up, and the monthly cap

Direct answer: Top-ups add prepaid credit. Credit pays for Cursor models beyond the weekly pool at published per-token rates, and for third-party models from the first request. Auto Top-up refills credit when your balance drops below a threshold you set.

Beyond-pool Cursor-model rates are published per million tokens. Grok 4.7 is $2 input and $6 output, and Composer 2.5 is $0.5 input and $2.5 output (Docs). A monthly additional usage limit caps paid additional usage each month, whatever credit you hold. When you hit your weekly limit, you can wait for the reset, use credit, or upgrade. Requests are not downgraded in quality or speed (Docs).

Third-party model billing

Direct answer: Claude, GPT, and Gemini models sit outside the pool. On Premium and above they are available at API list price from the first request, paid from prepaid credit. Free and Plus cannot use them.

Published list prices include Claude Sonnet 5.5 at $2 input and $10 output, GPT-5.6 Terra at $2 and $12, and Gemini 4 Carbon at $2 and $10, all per million tokens (Docs).

Legacy plans (short note)

Direct answer: Hobby, Start, Pro, Pro+, Ultra (legacy), and Teams Standard or Premium seats continue for existing customers. They use two monthly pools, Cursor Models and Other Models, plus on-demand usage instead of top-ups (Docs). If you already hold one, see the legacy plan documentation.

Comparison table

Direct answer: This table lists every published price and usage label across the plans. Usage labels are quoted as published, and no absolute request count appears because none is published.

PlanMonthlyAnnualPublished usageThird-party modelsStandout adds
Free$0—LimitedNot availableLimited agent and chat requests
Plus$8$80/yrBase (“Basic” in grid)Not availableGrok Chat, Voice, Imagine, Build, Grok API
Premium$30$300/yr4x PlusAPI list price, from prepaid creditCloud Agents, Automations, Origin, Bugbot, Grok Bot
Super$100$1,000/yr15x PlusAPI list price, from prepaid creditPriority model access at peak times
Ultra$200$2,000/yr50x PlusAPI list price, from prepaid creditEarliest access to all products and models
Teams$30 / $100 / $200 per seat—Matches the seat’s tier (4–50x Plus)API list price plus Cursor Token RateSSO, analytics, shared balance and cap
EnterpriseCustom—PooledAPI list price plus Cursor Token RateInvoicing, SCIM, spend controls, priority support

Sources: Pricing, Docs.

FAQ

Direct answer: The four questions below cover the points beginners ask about most, answered from the two official pages.

Is the weekly pool shared with Grok?

Yes. Grok Chat, Imagine, Voice, Grok Bot, and Cursor all draw from the same weekly pool for Cursor models, and usage rates vary by product (Docs). The Grok API is billed separately.

Can I use Claude, GPT, or Gemini on Plus?

No. Plus includes Cursor’s own models only. Upgrade to Premium or higher to use third-party models, billed at API list price from prepaid credit (Docs).

What happens when I hit my limit?

Wait for the weekly reset, use prepaid credit, or upgrade. The monthly additional-usage limit keeps paid spending bounded (Docs).

Is annual billing cheaper?

The published annual prices are $80 for Plus, $300 for Premium, $1,000 for Super, and $2,000 for Ultra, compared with $8, $30, $100, and $200 per month (Docs).

The Bottom Line

Direct answer: Cursor’s own guidance is Premium for new Agent users, Super for daily Agent users, Ultra for power users running several agents or Automations, and Plus for social creators with limited AI usage. Absolute numbers are not published, so budget with the monthly cap rather than guessing request counts.

Premium is the lowest tier that adds Cloud Agents, Automations, and third-party models, so it is the practical starting point for a newcomer (Pricing). Move up only when the weekly pool is the actual constraint. Subscriptions are sold only through the official site, prices exclude tax, and you should cancel at least 24 hours before renewal to avoid charges.

How This Guide Was Built

Direct answer: Every price, multiplier, and feature here comes from two pages fetched on 2026-11-15: Cursor’s public pricing page and its account pricing documentation. Nothing was inferred, estimated, or back-solved from published multipliers.

This review is based only on those two official pages. We did not run the tool hands-on.

Sources used throughout: cursor.com/pricing and cursor.com/docs/account/pricing.

Not verified

Direct answer: Four gaps remain. They are named here instead of being filled with unsourced numbers, and nothing in this article was observed in a live account.

  • Absolute limits not published: neither page gives per-tier request, credit, or token-pool counts.
  • Per-product draw rates: the docs say usage costs “vary by product” but publish no per-product draw rates.
  • Regional data residency: the 10% model-pricing uplift for eligible models is published, but eligible regions and models were not verified.
  • No hands-on testing: no paid or free account was run. Reset timing, notifications, and dashboard behaviour are as the docs describe, not observed.

📊 See how Cursor compares → /comparisons/

Back to all posts