How Cursor Pricing and Credits Actually Work (2026)

New customers buy Cursor through The X SubscriptionThe subscription Cursor is sold through: one plan for Grok, Cursor and X, with Free, Plus, Premium, Super and Ultra tiers, per-seat Teams, a weekly included pool for Cursor models shared with Grok, and third-party models billed at API list price from prepaid credit. Press Enter for the full definition., one subscription that covers Grok, Cursor and X. The plan ladder is Free, Plus, Premium, Super and Ultra for individuals, with Teams priced per seat and Enterprise on a custom contract. The usage column is relative to Plus, which is the unit Cursor uses to size every tier above it.

Cursor's own sizing advice is that Premium suits new Agent users, Super suits daily Agent users, and Ultra suits power users running several agents or Automations. Super also adds priority model access at peak times, and Ultra adds earliest access to new products and models.

Free covers Agent, Chat and Tab with limited usage on Cursor models. Teams are covered further down. If you bought Cursor before The X SubscriptionThe subscription Cursor is sold through: one plan for Grok, Cursor and X, with Free, Plus, Premium, Super and Ultra tiers, per-seat Teams, a weekly included pool for Cursor models shared with Grok, and third-party models billed at API list price from prepaid credit. Press Enter for the full definition. launched, you are probably on a legacy Cursor plan, covered below, and it keeps working exactly as it did.

Each paid plan includes one pool of usage for Cursor models, and it refreshes weekly, so one heavy day does not spend your month. The pool is shared across everything on the subscription: Agent in the editor, the CLI, Cloud AgentsAgents that run in a Cursor-managed virtual machine, check out the repo, do the work and open a pull request, then shut down, with no load on your laptop. Press Enter for the full definition., Grok BotAn early AI teammate product from SpaceXAI and Cursor that works across apps on a persistent cloud computer and asks for approval on selected actions; launch materials describe it as beta, but Cursor says Beta Services terms do not apply. Press Enter for the full definition. and the Grok apps. Your remaining usage and the next reset date show in editor settings and under Usage at grok.com/dashboard.

Different models draw from the pool at different rates. Cursor's table of Cursor models gives the per-million-token rates that apply when you buy usage beyond the pool, and they also show how fast each model drains it.

When the weekly pool runs out, Cursor does not switch your request to a slower or lower-quality model. You can wait for the reset, keep working from prepaid credit at the same per-model rates, or upgrade to a tier with a larger pool.

Claude, GPT, Gemini and other third-party models are not part of the included pool. On Premium and above you can use them at the model's API list price from the first request, paid from prepaid credit. Free and Plus do not include them.

The practical consequence is that the sticker price is the floor, not the bill, for anyone who prefers a Claude or GPT model. Claude Opus 5.5Anthropic's top Opus-tier model in Cursor at $4 per million input tokens and $20 per million output, 20% below Claude Opus 5, which it replaces. Press Enter for the full definition. lists at $4 per million input tokens and $20 per million output, so a single run that reads a million tokens and writes a hundred thousand costs $6 in credit before cache discounts. The same run on Composer 2.5The current Composer release, better at long-running tasks and at judging when a job needs a light touch versus deep work. Press Enter for the full definition. would draw down the weekly pool instead. Decide per task which pool you want to spend.

Auto follows the same split. Cost mode (classic Auto) bills as a Cursor-model request and draws from your weekly pool whichever model serves it. Balance and Intelligence exist only on Teams and Enterprise with Cursor Router on; they bill at the routed model's list price, and a route to a third-party model is paid from prepaid credit. Individual plans have Cost mode only.

Teams are priced per seat. Each member holds a Premium, Super or Ultra seat at the same price and the same weekly usage as the matching individual plan, and one team can mix tiers. Plus is not available for teams, and a Free seat carries no team allowance. Each paid seat has its own weekly pool, and usage does not transfer between members.

Beyond the weekly pools the team shares one prepaid balance that only admins can top up, with one monthly additional usage limit for the whole team. Third-party models are billed at public API list prices plus the Cursor Token Rate of $0.25 per million tokens. Cursor models, including Grok and ComposerCursor's own fast coding model, tuned for the editor and priced well below frontier models; the recommended day-to-day model for executing a plan. Press Enter for the full definition., and Cost-mode Auto requests are exempt from that rate. Individual plans never pay it.

Legacy Cursor plansThe plans Cursor sold before The X Subscription (Hobby, Start, Pro, Pro+, Ultra, and Teams Standard and Premium seats); they keep working for existing subscribers with two monthly usage pools and on-demand usage. Press Enter for the full definition. are the plans Cursor sold before The X SubscriptionThe subscription Cursor is sold through: one plan for Grok, Cursor and X, with Free, Plus, Premium, Super and Ultra tiers, per-seat Teams, a weekly included pool for Cursor models shared with Grok, and third-party models billed at API list price from prepaid credit. Press Enter for the full definition.: Hobby, Start, Pro, Pro+, Ultra, and Teams with Standard and Premium seats. They are not sold to new customers, and existing subscribers keep them until they switch or cancel. Switching to The X Subscription is optional and permanent.

The structure underneath is different. Legacy plans split included usage into two monthly pools: Cursor ModelsOn legacy Cursor plans, the monthly usage pool for Cursor's own models (Grok 4.7, Grok 4.6, Grok 4.5 and Composer 2.5), kept separate from the third-party Other Models pool. The X Subscription replaces it with one weekly pool. Press Enter for the full definition. covers Grok 4.7Cursor and SpaceXAI's current lead Cursor model, in the weekly pool at $2 per million input tokens and $6 per million output, with a 256k context window and a 500k variant; above 256k input tokens it bills at 2x the standard rate. Press Enter for the full definition., Grok 4.6A Cursor and SpaceXAI model for complex coding and knowledge work, still in the Cursor model pool and succeeded by Grok 4.7; it is a selectable model and is separate from the Grok Bot product. Press Enter for the full definition., Grok 4.5 and Composer 2.5The current Composer release, better at long-running tasks and at judging when a job needs a light touch versus deep work. Press Enter for the full definition., and Other ModelsOn legacy Cursor plans, the monthly pool for supported third-party models from providers such as Anthropic, Google and OpenAI, separate from Cursor Models. On The X Subscription third-party models are billed from prepaid credit instead. Press Enter for the full definition. covers third-party models at their API prices. Extra requests bill as on-demand usage in arrears. Cursor says each paid plan on The X SubscriptionThe subscription Cursor is sold through: one plan for Grok, Cursor and X, with Free, Plus, Premium, Super and Ultra tiers, per-seat Teams, a weekly included pool for Cursor models shared with Grok, and third-party models billed at API list price from prepaid credit. Press Enter for the full definition. includes at least twice as much coding usage with Cursor models as the legacy plan it replaces, and drops the included third-party allowance.

Cursor StartAn India-only legacy individual plan at ₹649 a month centered on Cursor's own models, with cloud and mobile access but without Other Models, Auto, Bugbot, Automations or the Agent SDK; Cursor names Plus as the comparable plan on The X Subscription. Press Enter for the full definition. is the India-only legacy plan at ₹649 a month, tax inclusive and billed in INR. It covers the Cursor ModelsOn legacy Cursor plans, the monthly usage pool for Cursor's own models (Grok 4.7, Grok 4.6, Grok 4.5 and Composer 2.5), kept separate from the third-party Other Models pool. The X Subscription replaces it with one weekly pool. Press Enter for the full definition. pool only (non-fast Grok 4.7Cursor and SpaceXAI's current lead Cursor model, in the weekly pool at $2 per million input tokens and $6 per million output, with a 256k context window and a 500k variant; above 256k input tokens it bills at 2x the standard rate. Press Enter for the full definition., Grok 4.6A Cursor and SpaceXAI model for complex coding and knowledge work, still in the Cursor model pool and succeeded by Grok 4.7; it is a selectable model and is separate from the Grok Bot product. Press Enter for the full definition., Grok 4.5 and Composer 2.5The current Composer release, better at long-running tasks and at judging when a job needs a light touch versus deep work. Press Enter for the full definition.), plus Cloud AgentsAgents that run in a Cursor-managed virtual machine, check out the repo, do the work and open a pull request, then shut down, with no load on your laptop. Press Enter for the full definition., Cursor for iOS, plugins, MCPModel Context Protocol. A standard that lets an AI agent pull in context from outside the repo, like Jira tickets or internal docs. Press Enter for the full definition. servers, hooks and skills. It has no on-demand option and excludes BugbotCursor's automated PR reviewer that posts inline findings and can push fix commits from isolated VMs. Press Enter for the full definition., Auto, Automations and the Cursor SDK. Max Mode exists only on the older request-based plans, where it extends a model beyond its default context at the model's API rate plus 20%.

You control Cursor cost in two ways: cap what you can spend, and lower what each task burns. Set a spending cap so the bill can never run away, then trim token use per request. These seven steps cover both, from the hard limit down to day-to-day model choices.

1. 1Set a monthly additional usage limit at grok.com/dashboard/billing (or an on-demand spend limit on a legacy plan) so you can never be surprised. Leave Auto Top-up off until you know your burn rate.
2. 2Use a smaller or cheaper model for routine edits; reserve the strongest models for hard problems.
3. 3Watch your weekly pool and your prepaid credit separately, and on Teams check which Cursor Router mode is the default.
4. 4Give the small group of heavy agent users Super or Ultra seats instead of upgrading every teammate by default.
5. 5Keep context tight: fewer open tabs and scoped @-context cost fewer tokens per request.
6. 6Batch related work into one agent session instead of many cold starts.
7. 7Watch the usage dashboard for the first month to learn your real burn rate.

For developers who code daily, Premium pays for itself on simple arithmetic: if your time is worth $30 an hour, it needs to save about an hour a month. The honest caveat is that the $30 buys a weekly pool of Cursor models, not Claude or GPT. If you lean on third-party models all day, you will pay for them from prepaid credit at API list price. Super ($100) or Ultra ($200) is the realistic cost only if the bigger weekly pool is what you run out of. Budget for your actual model mix rather than the sticker price.

There is no fixed per-request price to anchor to; usage is measured in model compute, not a request count. So picture the day instead. A couple of large agent runs on a frontier model, plus a long planning conversation, can quietly draw down a meaningful slice of a weekly pool, or of prepaid credit, in one sitting. That is why heavy users feel the burn well before the week ends: the spend tracks how much compute each prompt actually consumed, and the most capable models on long context consume the most.

The plan-then-execute cost loop

A reliable pattern keeps cost and quality in balance: plan with a frontier model, then execute with ComposerCursor's own fast coding model, tuned for the editor and priced well below frontier models; the recommended day-to-day model for executing a plan. Press Enter for the full definition.. You spend a little on the most capable model for a tight, reviewed plan, then hand the mechanical edits to the cheaper, fast Composer model. The expensive thinking happens once, up front, where it determines whether everything downstream is right. Most of the token volume runs at the cheaper rate.

1. 1Plan the change with a frontier model so the approach is correct before any code moves.
2. 2Review and tighten the plan: this is the cheapest place to catch a wrong direction.
3. 3Execute the plan with ComposerCursor's own fast coding model, tuned for the editor and priced well below frontier models; the recommended day-to-day model for executing a plan. Press Enter for the full definition., which carries the high token volume at a lower rate.
4. 4Reserve frontier-model turns for the genuinely hard reasoning, not the bulk edits.