Cursor Pricing After SpaceX: Plus, Premium, Super, Ultra and The X Subscription Explained (October 2026)
Cursor's pricing page changed completely in October. SpaceX closed its acquisition of Cursor on August 14, 2026. On October 12, SpaceXSI (the name now used on x.ai, whose footer credits SpaceXAI LLC) launched The X Subscription, a single plan that covers Grok, Grok Bot, Cursor and X Premium. The old Cursor tiers (Hobby, Pro, Pro+, Ultra, Teams Standard and Premium) are now legacy plans. New customers buy Cursor through X Subscription Corp, and their usage pool is shared with the Grok apps.
If you pay for Cursor seats, three things matter. Prices went up for individuals moving from Pro, and down for teams. Claude, GPT and Gemini usage inside Cursor is now pay-as-you-go on every plan that offers it (Premium and above), with no included allowance; Free and Plus cannot use third-party models at all. The usage allowance now refreshes weekly instead of monthly. Every price below was checked on October 26, 2026 against Cursor's pricing page and docs, the X Subscription announcement on x.ai, and competitors' pricing pages.
What changed: SpaceX, Cursor and The X Subscription
| Date (2026) | Event | Source |
|---|---|---|
| April | Cursor announces a partnership with SpaceXAI (Cursor's wording) to speed up model training. Cursor's blog later called this the start of the acquisition process. TechCrunch reported that the April deal let SpaceX either buy Cursor for $60B in stock or pay a $10B break-up fee. | Cursor blog, TechCrunch |
| June 16 | TechCrunch reports SpaceX will acquire Cursor for $60B in stock, days after SpaceX's IPO, with closing expected in Q3. | TechCrunch |
| August 14 | Cursor's blog post: "Cursor has officially been acquired by SpaceX." | Cursor blog |
| October 12 | The X Subscription launches, bundling Grok, Grok Bot, Cursor and X Premium. It opens to new customers that day and rolls out globally over the following days. | x.ai announcement (SpaceXSI) |
The practical effect is that Cursor's pricing page is now an X Subscription pricing page. It lists X features such as the checkmark, ad-free scrolling and X Money cashback next to Cursor's agent features. Subscriptions are sold under the SpaceXSI Subscription Terms. The pricing FAQ says plainly that subscriptions "are only sold directly through X Subscription Corp" and that resold accounts may be suspended. Billing has moved from cursor.com to grok.com/dashboard/billing.
Cursor plans and prices (as of October 2026)
| Plan | Monthly | Annual | Included usage | Cursor agent features | Third-party models | Grok Bot |
|---|---|---|---|---|---|---|
| Free | $0 | n/a | Limited | Agent and Chat, limited requests | Not available | No |
| Plus | $8 | $80/yr | Base (1x) | Agent with Cursor's own models only, fixed settings | Not available | No |
| Premium | $30 | $300/yr | 4x Plus | Adds cloud agents and automations, Origin code hosting, Bugbot code reviews | API list price, from prepaid credit | Yes |
| Super | $100 | $1,000/yr | 15x Plus | Premium, plus priority model access at peak times | API list price, from prepaid credit | Yes |
| Ultra | $200 | $2,000/yr | 50x Plus | Super, plus earliest access to new products and models | API list price, from prepaid credit | Yes |
| Teams | $30 / $100 / $200 per seat | Not listed | Matches the seat tier | Premium features, plus team marketplace, admin, SAML/OIDC SSO | API list price plus $0.25 per 1M token Cursor Token Rate on eligible models | Yes |
| Enterprise | Custom | Custom | Pooled | Teams, plus SCIM, audit logs, invoicing, granular controls | API list price plus Cursor Token Rate | Yes, with access and audit controls |
Prices exclude tax. The annual prices are on Cursor's models-and-pricing docs page, not the main pricing page. Annual billing works out to two months free on each paid tier. Plus costs $96 a year billed monthly against $80 annually, which saves $16. Premium is $360 against $300 (saves $60), Super is $1,200 against $1,000 (saves $200), and Ultra is $2,400 against $2,000 (saves $400).
What each tier actually includes
Free ($0)
Limited Agent and Chat requests in Cursor, limited Grok Chat and image generation, and standard X. The x.ai announcement calls it "a starter allowance for Cursor's coding agents." It works for evaluation. It will not cover daily work.
Plus ($8): Cursor's own models only
Plus is the new entry tier and has the most restrictions. Cursor's docs describe it as Agent with Cursor's own models (Grok 4.7, Grok 4.6, Grok 4.5 and Composer 2.5) "at fixed settings: Fast is off, effort is medium, and Grok 4.7 uses the 256k context window." The same docs list what Plus leaves out: "No Auto, Cloud Agents, Automations, Bugbot, Origin, third-party models, or your own API keys." The pricing card marks it "*Only Grok models are available on this plan," but the docs also list Composer 2.5, Cursor's own model, for Plus. Plus also cannot be bought as a team seat. Cursor's own FAQ recommends Plus for "social creators with limited AI usage," not for developers.
Premium ($30): the real developer tier
Premium is labeled "Most popular" and is where Cursor recommends new AI users start. On top of Plus it adds 4x the usage, the latest Grok models without the fixed settings, cloud agents and automations, code hosting via Origin, AI code reviews via Bugbot, third-party models at API list price, Grok Bot, and video generation. The x.ai announcement says Premium bundles "the benefits of SuperGrok, Cursor Pro, and X Premium+, previously $90/month combined." For an engineer, Premium is the first tier that matches what the old Cursor Pro offered. If you are new to the editor, our Cursor 3 and Composer walkthrough covers the agent features available on these tiers.
Super ($100) and Ultra ($200)
Super adds 15x Plus usage and priority model access at peak times. Cursor recommends it for daily agent users. Ultra adds 50x Plus usage and earliest access to new products and models, and Cursor pitches it at power users "running several agents or Automations." Neither tier includes a third-party model allowance. Spending more gets you a larger Grok and Composer pool, not more Claude or GPT tokens.
Usage per dollar: Ultra is the best deal if you use it
Because the multiples grow faster than the prices, the included pool costs less per unit at each higher tier. Taking Plus usage as 1 unit:
| Tier | Price | Usage units | Units per $ | vs Premium |
|---|---|---|---|---|
| Plus | $8 | 1 | 1 / 8 = 0.125 | 0.94x |
| Premium | $30 | 4 | 4 / 30 = 0.133 | 1.00x |
| Super | $100 | 15 | 15 / 100 = 0.150 | 1.13x |
| Ultra | $200 | 50 | 50 / 200 = 0.250 | 1.88x |
Cursor's team docs give the same ratios: a Super seat has 3.75x the Premium pool (15 / 4) and an Ultra seat has 12.5x (50 / 4). Ultra is only good value if your team actually uses that much each week. An unused weekly pool does not carry over.
The third-party model question: the table vs the FAQ
This part of the pricing page reads inconsistently, so here is the exact wording rather than a paraphrase. As of October 26, 2026, the Premium card on cursor.com/pricing lists, under "Everything in Plus, and:":
"Third-party models at API list price"
The "Compare features across plans" table puts that same line in the Cursor section with a checkmark for Premium, Super, Ultra, Teams and Enterprise, next to features that are included in the price, such as cloud agents and Bugbot. The FAQ entry "What models are supported?" says:
"Plus includes Grok models only. Premium and above include generous usage of Grok models but do not include an allowance for third-party models. Third-party models remain available in Cursor at API list price. You pay for exactly what you use, and you can set spending caps to manage costs."
The fine print below the cards says: "Third Party model usage in Cursor is not included in the monthly prices shown and must be purchased separately."
Taken together, the FAQ, the fine print and the docs agree with each other. The checkmark means you are allowed to use Claude, GPT or Gemini in Cursor. It does not mean any usage is included. Cursor's docs say third-party models "aren't part of the included pool" and are charged "from the first request, paid from your prepaid credit." If you are moving from legacy Pro, which had an "Other Models" pool, this is the biggest change. The legacy-plans page says the new tiers drop "the included third-party model allowance."
What third-party usage costs
Cursor publishes a per-model price list. A few common coding models, per million tokens, as listed in Cursor's docs on October 26, 2026:
| Model | Input | Cache write | Cache read | Output |
|---|---|---|---|---|
| Claude Opus 5.5 | $4 | $5 | $0.20 | $20 |
| Claude Sonnet 5.5 | $2 | $2.50 | $0.10 | $10 |
| GPT-5.6 Sol | $4 | $5 | $0.40 | $20 |
| GPT-5.6 Terra | $2 | $2.50 | $0.20 | $12 |
| Gemini 3.1 Pro | $2 | n/a | $0.20 | $12 |
| Grok 4.7 (beyond the weekly pool) | $2 | n/a | $0.50 | $6 |
| Composer 2.5 (beyond the weekly pool) | $0.50 | n/a | $0.20 | $2.50 |
The docs add that opting in to regional data residency adds a 10% uplift on eligible models. For how Grok 4.7 API pricing compares with these subscription rates, see our Grok 4.7 pricing and API cost breakdown. Grok API calls made outside Cursor are billed separately and do not draw from the subscription pool.
How the weekly shared pool, top-ups and caps work
- One pool, shared with Grok. Each paid plan has a single pool for Cursor models (Grok 4.7, 4.6, 4.5 and Composer 2.5). Agent in the editor, the CLI, cloud agents, Grok Bot and the Grok apps (Chat, Imagine, Voice) all draw from it. Usage rates vary by product. Heavy Grok Imagine use in the evening means less agent time the next morning.
- Weekly refresh. Cursor's FAQ: "Your allowance refreshes weekly, so one heavy day never wipes out your whole month." Your remaining usage and reset date appear in editor settings and at grok.com/dashboard.
- When it runs out. You can wait for the reset, keep going on prepaid credit ("top-offs" on the pricing page, "top-ups" in the docs), or upgrade. Cursor says requests "are never downgraded in quality or speed." Pool overage is billed at the per-model rates above.
- Auto Top-up and caps. Auto Top-up refills credit when the balance drops below a threshold you set. A monthly additional usage limit caps paid usage no matter how much credit you hold. Set a cap before you turn on Auto Top-up.
- Auto mode. Individual plans get only Cost mode (classic Auto), which bills as a Cursor-model request against the pool. The Balance and Intelligence router modes, which can route to third-party models, are only on Teams and Enterprise.
Grok Bot is included from Premium up and draws from the same pool. Our Grok Bot guide covers what it does and how it fits next to Cursor's cloud agents.
Cursor Teams and Enterprise pricing
| Seat | Price per user/month | Weekly pool |
|---|---|---|
| Premium | $30 | Premium pool (4x Plus) |
| Super | $100 | 3.75x Premium |
| Ultra | $200 | 12.5x Premium |
| Free | $0 | No team allowance |
- Mix and match. One team can combine seat tiers, for example Premium for most engineers and Super for the people running agents all day. Plus is not available for teams. Mid-cycle seat additions and upgrades are prorated, and downgrades take effect at the end of the billing period.
- Pools stay per seat. Each paid seat has its own weekly pool, and unused usage does not move between seats. Pooled usage is an Enterprise feature.
- Shared credit and caps. Third-party usage and post-pool usage draw from one team prepaid balance. Admins set a team-wide monthly limit, a default per-member limit, and per-member overrides.
- Cursor Token Rate. On Teams and Enterprise, eligible third-party requests carry an extra $0.25 per million tokens on input, output and cached tokens, including BYOK usage. Cursor's own models, Cost mode, and a list of small models (Claude Haiku 5.5, GPT-5.4 Mini, Gemini 3.8 Flash and others) are exempt. Individual plans do not pay it.
- Admin and security. Teams includes centralized billing, usage analytics, team-wide privacy settings, custom Chat data retention and SAML/OIDC SSO. The x.ai announcement also lists SOC 2 compliance for teams. Cursor's site footer lists SOC 2, ISO 27001, ISO 42001 and AIUC-1. Enterprise adds invoice/PO billing, SCIM, granular and group-level spend controls, repository, model and MCP access controls, audit logs, service accounts, Grok Bot audit controls and priority support.
The x.ai announcement says the new team seats are "priced lower per seat than older Cursor Teams but now includes 3x more coding usage." The legacy seat prices below show the reduction.
Legacy plans: keep or switch?
Nobody is forced to move. Cursor's docs: "Your legacy plan, price, and features continue unless you change or cancel." No end date for legacy plans has been published. Switching is optional but permanent. You cannot go back.
| Legacy plan | Legacy price | Comparable X Subscription tier | Monthly change |
|---|---|---|---|
| Hobby | Free | Free | $0 |
| Pro | $20 | Premium ($30) | +$10 |
| Pro+ | $60 | Super ($100) | +$40 |
| Ultra | $200 | Ultra ($200) | $0 |
| Teams Standard | $40/user | Teams Premium seat ($30) | -$10 per seat |
| Teams Premium | $120/user | Teams Super seat ($100) | -$20 per seat |
| Start (India only) | ₹649, tax inclusive | Plus ($8), as an analogy only | n/a |
Start is a legacy India-only plan that can only be bought from a free account. Cursor's legacy-plans docs call the Start-to-Plus mapping "an analogy only" and say that "when a paid Cursor account switches, the unify flow offers Premium and above." Cursor says each new paid tier includes "at least twice as much coding usage with Cursor models as the legacy plan it replaces." The trade-offs:
- You gain: at least 2x Grok/Composer usage, Grok and X benefits, Grok Bot, and a weekly refresh.
- You lose: the included "Other Models" pool for third-party models, and monthly on-demand billing (replaced by prepaid top-ups). Max Mode is also only available on legacy request-based plans.
- Refunds: eligible Cursor and Grok subscriptions are cancelled at checkout and refunded pro rata for unused time. Cursor subscriptions billed through the Apple App Store have to be cancelled by you in iOS Settings. Any uninvoiced on-demand usage is billed in a final invoice before the switch completes. Limited-time switching offers appear in billing settings when they are available.
Rule of thumb: if most of your Cursor usage runs on Claude or GPT, a legacy Pro or Teams plan with its included Other Models pool may cost less than switching. If your team already defaults to Grok 4.7 or Composer, switching gets you more usage, and team switches also lower the seat price.
Cost per developer: Cursor vs Claude Max vs Copilot
Seat prices alone are misleading, because the new Cursor tiers charge for third-party tokens separately. The scenarios below use list prices as of October 2026. The token volumes are an assumption for illustration: 8M uncached input, 4M cache-write, 30M cache-read and 2M output tokens per developer per month of third-party model use. Your real numbers depend on how much agent work you run. Check your Cursor usage dashboard and substitute your own figures.
Scenario 1: solo developer, Grok and Composer only
Premium at $30/month, or $300/year ($25/month equivalent). If the weekly pool covers your work, that is the full cost.
Scenario 2: solo developer who wants Claude in Cursor
Claude Sonnet 5.5 at Cursor's listed rates ($2 input, $2.50 cache write, $0.10 cache read, $10 output per 1M tokens):
- Input: 8M x $2 = $16.00
- Cache write: 4M x $2.50 = $10.00
- Cache read: 30M x $0.10 = $3.00
- Output: 2M x $10 = $20.00
- Third-party total: $49.00, so Premium + Sonnet = $30 + $49 = $79/month
The same volume on Claude Opus 5.5 ($4 / $5 / $0.20 / $20): 32 + 20 + 6 + 40 = $98, so Premium + Opus = $128/month.
For comparison, Anthropic lists Claude Max at $100/month (5x Pro usage) and $200/month (20x Pro usage). Both include Claude Code under rolling five-hour and weekly limits instead of per-token billing. Claude Pro is $20/month, or $17/month billed annually. A developer who spends most of the day in Claude models will usually get more predictable costs from Max with Claude Code than from paying API list price inside Cursor. Our Cursor vs Claude Code vs Codex vs Gemini CLI comparison covers the workflow differences.
Scenario 3: 10-engineer team on Cursor Teams
Ten Premium seats with the Scenario 2 Sonnet usage per engineer:
- Seats: 10 x $30 = $300
- Sonnet tokens at list price: 10 x $49 = $490
- Cursor Token Rate: (8 + 4 + 30 + 2) = 44M tokens x $0.25 = $11 per engineer; 10 x $11 = $110
- Total: $300 + $490 + $110 = $900/month, or $90 per engineer
A mixed team of 6 Premium, 3 Super and 1 Ultra seats costs 6 x $30 + 3 x $100 + 1 x $200 = $180 + $300 + $200 = $680/month in seats before any third-party usage.
How that compares (seat prices, as of October 2026)
| Option | Per user/month | 10 users/month | What usage is included |
|---|---|---|---|
| Cursor Teams, Premium seats | $30 | $300 | Weekly Grok/Composer pool; third-party models pay-as-you-go plus Token Rate |
| Cursor Teams, with Scenario 3 Sonnet usage | $90 | $900 | As above, including the illustrative third-party spend |
| Claude Team, Premium seat | $125 monthly / $100 annual | $1,250 / $1,000 | Claude Code included on the Team plan; 5x Standard seat usage, subject to limits |
| Claude Max 5x (individual) | $100 | n/a (individual plan) | 5x Pro usage, Claude Code included |
| GitHub Copilot Business | $19 | $190 | 1,900 GitHub AI Credits per user per month |
| GitHub Copilot Enterprise | $39 | $390 | 3,900 GitHub AI Credits per user per month |
GitHub's plans page values one AI Credit at $0.01, so Business includes about $19 of model usage per seat and Enterprise about $39. Individual Copilot plans are Pro ($10), Pro+ ($39) and Max ($100). These products differ, so treat the table as a budgeting baseline rather than a like-for-like comparison. Among agent-first tools, Cursor's $30 seat is the cheapest way in, provided Grok and Composer cover most of your work. Copilot Business is cheaper per seat but includes far less model usage. Once you add heavy Claude or GPT usage, Cursor Teams lands close to Claude's Team pricing.
Which Cursor plan should you pick?
- Trying Cursor: Free. Plus is cheap but has no cloud agents, Auto, Bugbot or third-party models, so it is a poor test of the editor.
- Individual engineer, mostly Grok/Composer: Premium, paid annually if you are sure you will keep it.
- Daily agent user hitting the weekly reset: Super. Look at your usage dashboard before going to Ultra. Ultra has the best usage per dollar, but only if you use the extra pool.
- Claude-first engineer: stay on legacy Pro if you have it, or compare Premium plus per-token spend against Claude Max.
- Team of 5-50: Teams with mostly Premium seats and a few Super seats, a team-wide monthly cap, and per-member limits configured from day one.
- Regulated or large org: Enterprise for SCIM, audit logs, model and MCP access controls, and pooled usage.
For how Cursor compares with the rest of the field, see our top AI coding tools for 2026. If you are hiring for these workflows, our guide to hiring AI-native engineers who use Claude Code and Cursor covers what to screen for.
FAQ
How much does Cursor cost in 2026?
As of October 2026, Free is $0, Plus $8/month, Premium $30, Super $100 and Ultra $200. Annual billing is $80, $300, $1,000 and $2,000 a year. Teams seats are $30, $100 or $200 per user per month, and Enterprise is custom.
Is Cursor Pro still available?
Not to new customers. Pro ($20), Pro+ ($60), Ultra ($200) and the old Teams seats continue for existing subscribers as legacy plans unless they switch or cancel. Pro's closest new tier is Premium ($30).
Does Cursor Premium include Claude?
Premium gives you access to Claude and other third-party models, but no included allowance. Every request is billed at API list price from prepaid credit, for example $2 input and $10 output per 1M tokens for Claude Sonnet 5.5.
What is the difference between Cursor Premium and Ultra?
Both include the same core Cursor agent features: cloud agents and automations, Bugbot, Origin and third-party models at API list price. Ultra adds 50x Plus usage against Premium's 4x (12.5x Premium's pool), priority model access at peak times and earliest access to new products and models. Ultra costs $200 against Premium's $30.
Does my Grok usage count against my Cursor usage?
Yes. Grok Chat, Imagine, Voice, Grok Bot and Cursor share one weekly pool. Grok API usage is billed separately and does not draw from it.
Do I get a refund if I switch from a legacy plan?
Yes. Eligible Cursor and Grok subscriptions are cancelled at checkout with a prorated refund for unused time. App Store-billed Cursor subscriptions must be cancelled manually in iOS Settings.
Sources
- Cursor pricing (X Subscription plans and FAQ), checked October 26, 2026
- Cursor docs: Models and pricing
- Cursor docs: Team pricing
- Cursor docs: Legacy Cursor plans
- Cursor blog: Cursor is now a part of SpaceX (Aug 14, 2026)
- SpaceXSI (x.ai): Introducing The X Subscription (Oct 12, 2026)
- TechCrunch: SpaceX to acquire Cursor for $60B in stock
- Yahoo Finance: SpaceX bundles Grok, X and Cursor
- Claude pricing (Anthropic)
- Anthropic Help Center: What is the Max plan?
- GitHub Docs: Plans for GitHub Copilot
- GitHub Copilot plans
If you need engineers who can make good use of whichever plan you pick, Codersera can help you extend your engineering team with vetted remote developers who are already fluent in Cursor, Claude Code and other AI coding tools.