Prices checked November 9, 2026 · Desk review, no paid plan

Cursor Pricing 2026: What Each Plan Really Costs You

Five individual tiers, a shared weekly usage pool and a third-party model bill on top. Here is the maths, the limits and who should skip it.

Summary

Cursor pricing changed on October 12, 2026: new subscribers now buy The X Subscription, with Free, Plus at $8, Premium at $30, Super at $100 and Ultra at $200 a month. Claude, GPT and Gemini are no longer included on any tier and are billed at API list price from Premium up. Great value if you code on Cursor's own models, much harder to forecast if Claude is your daily driver.

6.3 /10

Cursor is still one of the strongest AI editors, but the 2026 X Subscription plans changed the maths. Premium at $30 a month is the first plan with Claude, GPT and cloud agents, yet third-party models are billed on top at API list price. We score it 6.3/10: good value on Cursor's own models, expensive and hard to forecast if Claude is your daily driver.

Agent and editor quality
8.5/10
Ease of use
8.5/10
Value on Cursor's own models
8/10
Value with Claude or GPT
4.5/10
Price clarity
4.5/10
Billing and support trust
4/10
  • Annual billing takes 16.7% off every paid tier, two months free on each
  • Price per unit of usage falls from $7.50 on Premium to $4.00 on Ultra
  • Agent, cloud agents and Bugbot are mature: 4.6/5 across 332 G2 reviews
  • Claude, GPT and Gemini are billed at API list price on top of every plan
  • The weekly usage pool is published only as multiples of Plus, never in dollars or tokens
  • Billing and refund complaints dominate Trustpilot, where Cursor holds 1.5/5

Prices exclude tax. Official site, no affiliate link

Methodology

How we audited the pricing

Plan paid
None. No paid subscription was purchased for this review
Version tested
Cursor pricing page and docs as of November 9, 2026 (The X Subscription plans, Cursor 3.29 download build)
Test period
2026-11-09 → 2026-11-09
Test categories: Plan structure • Third-party model rates • Legacy plan migration • Cost modelling • Customer review aggregation

This is a documentary audit, not a hands-on test. We read the live pricing page, the Models and Pricing docs and the legacy plans page on November 9, 2026, captured screenshots of those pages, and cross-checked the numbers against two independent write-ups. Ratings come from the public G2, Capterra, Trustpilot and Product Hunt profiles plus Reddit thread titles. We did not benchmark Cursor or measure the weekly usage pool, because Cursor does not publish its size. Every cost example is arithmetic on published list prices, labelled as such, and caching discounts are ignored.

Should you pay for Cursor?

YES if you...

  • Devs who live in the editor all day and are happy with Composer 2.5 or Grok 4.7 as the default model
  • Existing legacy Pro subscribers who mostly use Auto, since Premium gives at least twice the in-house usage

NO if you...

  • Anyone whose daily driver is Claude or GPT: every request is billed on top at API list price
  • Freelancers who must forecast the monthly bill to the dollar, given an unpublished weekly allowance
  • Legacy Pro and Pro+ users who rely on the included third-party model pool, since the new tiers drop it
Pricing

Individual plans on The X Subscription

Monthly prices in USD, before tax. Annual prices checked on the Models and Pricing docs page.

Free

$0 /mo

To try the editor

  • Limited Agent and Chat requests
  • No credit card required
  • No third-party models

Plus

$8 /mo

$80 a year

  • Grok 4.7, 4.6, 4.5 and Composer 2.5 only
  • Fixed settings, Grok 4.7 at 256k context
  • No Auto, Cloud Agents, Automations or Bugbot
  • No third-party models and no own API keys

Super

$100 /mo

$1,000 a year, 15x Plus usage

  • Everything in Premium
  • Priority model access at peak times
  • Third-party models still billed on top

Ultra

$200 /mo

$2,000 a year, 50x Plus usage

  • Everything in Super
  • Earliest access to new products and models
  • Third-party models still billed on top

ROI breakdown: Per unit of Plus-sized usage, Plus costs $8.00, Premium $7.50, Super $6.67 and Ultra $4.00.

Hidden costs & gotchas
  • Third-party models are paid from prepaid credit at API list price from the first request
  • Teams add a $0.25 per million token Cursor Token Rate on eligible third-party models
  • Regional data residency adds 10% to eligible model prices
  • Annual plans are prepaid for twelve months, and Trustpilot reviewers report no refunds for unused months
  • The weekly pool is shared with Grok Chat, Imagine, Voice and Grok Bot, not only with the editor

What reviewers say, by platform

Ratings read from public profiles on November 9, 2026. Capterra shows a very small sample. Trustpilot profile is unclaimed.

G2

4.6 / 5

332 reviews

Capterra

4.0 / 5

5 reviews

Trustpilot

1.5 / 5

356 reviews

Product Hunt

5.0 / 5

981 reviews

Audit

What we measured

Every figure is read from Cursor's published prices or computed from them. Nothing here comes from a benchmark run.

Entry price for third-party models
30 $ per month (Premium) Plus at $8 has no third-party models and no own API keys. Source: cursor.com/docs/models-and-pricing
Price per Plus-sized unit of usage
7.50 / 6.67 / 4.00 $ on Premium / Super / Ultra Monthly price divided by the published multiple (4x, 15x, 50x of Plus)
Annual billing discount
16.7 % on every paid tier $80, $300, $1,000 and $2,000 a year versus 12 monthly payments of $8, $30, $100 and $200
Legacy Pro to Premium price change
+50 % $20 to $30 a month. Pro+ to Super is +66.7% ($60 to $100). Ultra stays at $200
Opus 5.5 versus Composer 2.5 list price
8 x on input and on output $4 and $20 versus $0.50 and $2.50 per million tokens
Sample agent session, 1M input and 0.2M output tokens
1.00 / 4.00 / 8.00 $ on Composer 2.5 / Sonnet 5.5 / Opus 5.5 List-price arithmetic, caching ignored
Review score spread
3.1 points between G2 (4.6) and Trustpilot (1.5) Profiles read on November 9, 2026
Audit question: what does the Plus plan include and exclude?
Plus covers Cursor's own models (Grok 4.7, 4.6, 4.5 and Composer 2.5) at fixed settings, plus Grok Chat, Voice, Imagine, Build and Grok API access. It excludes Auto, Cloud Agents, Automations, Bugbot, Origin, third-party models and your own API keys.
Cursor docs Models and Pricing page with the plans table
Audit question: which tier is promoted, and what does it add over Plus?
The pricing page flags Premium at $30 as Most popular: 4x the usage of Plus, plus cloud agents and automations, code hosting, AI code reviews and third-party models at API list price.
Cursor pricing page with Free, Plus, Premium and Super plans
Audit question: what is the product Cursor sells beyond the editor?
The product page frames Cursor as agents that plan, write and review code across desktop, CLI, web and mobile.
Cursor product page showing the agents desktop interface with a plan open
Pros and cons

Pros and cons

Pros

  • A mature agent, cloud agents and PR review in one product G2 reviewers (4.6/5 over 332 reviews) keep praising ease of use, coding assistance and codebase context.
  • Cheap entry on Cursor's own models Plus at $8 and annual billing at 16.7% off make Cursor's in-house models the low-cost way in.
  • Usage gets cheaper per unit as you move up $7.50 per Plus-sized unit on Premium, $6.67 on Super and $4.00 on Ultra, based on the published multiples.

Cons

  • Third-party models are never included, on any tier Claude, GPT and Gemini are paid from prepaid credit at API list price.
  • The weekly allowance cannot be sized in advance Cursor publishes multiples of Plus, shared with Grok, and no dollar or token figure.
  • Billing trust is the weak point Trustpilot reviewers describe unexpected charges, annual plans without refunds and AI-first support.
Verdict

Final verdict

6.3 /10

Cursor is a good editor sold through a pricing model you have to read twice. If you code mostly on Composer 2.5 or Grok 4.7, Premium at $30 is fair and Plus at $8 is a cheap way to try an agent. Heavy users get more per dollar on Super and Ultra.

If Claude or GPT is your default, the maths turns against it. A single agent session of 1M input and 0.2M output tokens costs about $4.00 on Sonnet 5.5 and $8.00 on Opus 5.5 at list price, against $1.00 on Composer 2.5, and none of it comes out of the plan. Eight Sonnet-sized sessions add $32 to a $30 subscription.

Our 6.3/10 averages 8.5 for the agent and the interface, 4.5 for price clarity and 4.0 for billing and support trust. If you already have a legacy Pro and use third-party models, keep it. If you are new, start on Free, set a spending cap before any top-up, and measure your own weekly burn.

Agent and editor qualityEase of useValue on Cursor's own modelsValue with Claude or GPTPrice clarityBilling and support trust
  • Agent and editor quality 8.5/10 Agent, cloud agents and Bugbot are mature
  • Ease of use 8.5/10 A VS Code fork, so extensions and keybindings carry over
  • Value on Cursor's own models 8/10 Plus at $8 and annual billing at 16.7% off
  • Value with Claude or GPT 4.5/10 Billed at API list price on top of the plan
  • Price clarity 4.5/10 Allowance published only as multiples of Plus
  • Billing and support trust 4/10 G2 4.6/5 versus Trustpilot 1.5/5

Common questions

How much does Cursor cost per month in 2026?
For individuals, Free is $0, Plus $8, Premium $30, Super $100 and Ultra $200 a month before tax. Annual billing is $80, $300, $1,000 and $2,000 for Plus, Premium, Super and Ultra.
Is Cursor free?
Yes, there is a Free tier with limited Agent and Chat requests and no credit card. Third-party models such as Claude and GPT are not available on it.
Are Claude and GPT included in Cursor plans?
Not on the new plans. Third-party models are billed at API list price from prepaid credit, and only from Premium up. For example Claude Sonnet 5.5 lists at $2 per million input tokens and $10 per million output tokens.
What happened to Cursor Pro at $20?
Pro, Pro+, Ultra and Teams Standard and Premium are now legacy plans. They keep working for existing customers, with two monthly usage pools. Cursor maps Pro to Premium at $30 and Pro+ to Super at $100. Switching is optional and permanent.
How does the weekly usage limit work?
Each paid plan has one pool for Cursor models, shared with Grok Chat, Imagine, Voice and Grok Bot, refreshed weekly. When it runs out you can wait, buy a top-up or move to a higher tier.
Is annual billing worth it?
It saves 16.7% on every paid tier, two months of the monthly price. The catch is that Trustpilot reviewers report no refunds on unused months, so pay monthly for one cycle first to see where your weekly usage lands.
Is Cursor cheaper than GitHub Copilot or Claude Code?
If you want Claude, usually not. Hack'celeration lists Copilot Pro at $10 with $15 of AI credits and Claude Pro at $20 with Claude Code included. Cursor Premium costs $30 and bills Claude on top.

Update log

  1. Initial publication. Pricing page, Models and Pricing docs, legacy plans page and review profiles on G2, Capterra, Trustpilot and Product Hunt checked on November 9, 2026.