aiapicostcalcUnit economicsPricesBrandsCompareBenchmarksSkills & PluginsFind my modelGuidesChanges

Cursor token cost: what one request really costs

Short answer: Cursor burns about 50K input tokens (60% cached) and 3,500 output tokens per typical task, so one task costs $0.0760 on Grok 4.7. At 200 tasks a month that is $15.20; without caching the same task would cost $0.1210. Evidence: estimated.

Last verified 2026-11-18 · Sources: src/tools.json (Cursor token basis, estimated) and src/prices.json (rates) · Report an error

By Shangyi Duan · Tool token basis: Estimated

Tool token data: src/tools.json (reviewed 2026-11-05) · Model rates: src/prices.json (checked 2026-11-18) · Evidence: Estimated

Cost per task, three price points

One row per scenario from our tool data. The three price columns price the identical token counts on the cheapest, mid and premium reference models, so the difference is the model rate, not the workload.

ScenarioInput tokensCached inputOutput tokensCheapest current modelMid-tier reference modelPremium reference model
Light task
one short request with a small context and one or two tool calls
12K40%1,200$0.00137$0.0240$0.2736
Typical task
one coding or agent task with several tool calls and a medium context
50K60%3,500$0.00405$0.0760$0.8100
Heavy session
a multi-turn agent session with a large repo or long history; input counts all tokens re-sent across turns
200K75%10K$0.0115$0.2350$2.30

Rates: input $2 / cached $0.5 / output $6 per 1M on Grok 4.7. Prices checked 2026-11-18.

Monthly budget

Scenario at 200 tasks / monthCheapest current modelMid-tier reference modelPremium reference model
Light task$0.2736$4.80$54.72
Typical task$0.8100$15.20$162
Heavy session$2.30$47.00$460

At 200 tasks a month, one task at the typical size costs $0.0760 on Grok 4.7, so $100 covers about 1316 tasks. The same budget buys 123 tasks on the premium reference and 24691 on the cheapest one.

Recalculate this page in your browser

Server-rendered from the typical task on Grok 4.7 at 200 tasks a month. Change a field to recalculate; nothing is sent anywhere and the formula is the one under Method.

Cost per task
$0.0760
Cost per 1,000 tasks
$76.00
Monthly
$15.20

Where the tokens go

The provider basis behind these numbers

Cursor docs say third-party models are billed at API list price per token, Cursor's own models draw from a weekly usage pool, and Teams/Enterprise add a $0.25 per million token Cursor Token Rate. Cursor does not publish tokens per agent task, so tokens are modelled: rules files and indexed code snippets are injected, and each agent step re-sends the conversation plus tool results. The model values are cross-checked against measured coding-agent samples (Aider benchmark: about 12k prompt tokens per one-shot task; Claude Code session diagnostics: cache-heavy multi-turn sessions).

What caching is worth here

Our data does not assume a cached prefix for this tool (60% cached share), so the task price above is already the uncached input price. If the tool does repeat a stable prefix, adding it to the provider cache cuts the input side to $0.5 per 1M on Grok 4.7.

How to reduce it

Each tip changes a different part of the formula: caching changes the input rate, trimming history changes the input count, and routing simple steps to a cheaper model changes the rate for part of the work.

Sources and evidence

SourceRead onWhat it establishes
cursor.com2026-11-04Cursor plan pools, per-model input/cache/output rates and the Cursor Token Rate.
cursor.com2026-11-04How included usage, prepaid credit and on-demand usage reset and are billed.

Evidence label: Estimated

Cursor docs say third-party models are billed at API list price per token, Cursor's own models draw from a weekly usage pool, and Teams/Enterprise add a $0.25 per million token Cursor Token Rate. Cursor does not publish tokens per agent task, so tokens are modelled: rules files and indexed code snippets are injected, and each agent step re-sends the conversation plus tool results. The model values are cross-checked against measured coding-agent samples (Aider benchmark: about 12k prompt tokens per one-shot task; Claude Code session diagnostics: cache-heavy multi-turn sessions).

Method

The formula

Cost per task = (uncached input × input rate + cached input × cached rate + output × output rate) ÷ 1,000,000. Rates are the standard list prices in the price table (checked 2026-11-18); token counts come from src/tools.json (tool data reviewed 2026-11-05).

Why three reference models

The columns are the cheapest, median and most expensive current models by the cost of one typical agent task, named above so you can check each rate. They are a range, not a prediction of which model Cursor uses.

What the monthly figure assumes

200 tasks a month (the tasks-per-month field in our tool data), the same count for every scenario, at the standard rate with no batch discount. A heavy session can count as several tasks, so replace the number with your own before you put it in a budget.

Official links

Related skills & plugins

Other entries in the AI coding agents category. Each name links to its token-cost page or its official site.

Related models

The three reference models in the cost table, with their full price tiers.

More

Related

Questions

How much does Cursor cost per task?

One typical task costs $0.0760 on Grok 4.7, $0.00405 on the cheapest reference model (GPT-6 Luna) and $0.8100 on the premium reference (GPT-6 Cyber). The spread is the model rate: the token counts are identical in all three columns.

What does Cursor cost per month at 200 tasks?

$15.20 a month on Grok 4.7 at the typical task size. Without the 60% cached input assumption it would be $24.20. Per 1,000 tasks the token bill is $76.00.

Why does Cursor use so many input tokens?

Cursor's agent reads rules files, retrieves codebase chunks and then re-sends that context with each tool call, so cost scales with how much of the repo it loads and how many steps it takes. Requests routed to Cursor's own models consume the plan pool rather than being token-priced, which makes the same task cheaper on a paid plan and token-priced with third-party models. The token shape is 50K input tokens (60% cached) against 3,500 output per typical task; input is resent every step, output is not.

Are the Cursor token numbers measured or estimated?

The evidence label is "Estimated". Cursor docs say third-party models are billed at API list price per token, Cursor's own models draw from a weekly usage pool, and Teams/Enterprise add a $0.25 per million token Cursor Token Rate. Cursor does not publish tokens per agent task, so tokens are modelled: rules files and indexed code snippets are injected, and each agent step re-sends the conversation plus tool results. The model values are cross-checked against measured coding-agent samples (Aider benchmark: about 12k prompt tokens per one-shot task; Claude Code session diagnostics: cache-heavy multi-turn sessions). Sources and read dates are in the table above; if a number is an estimate, the basis says what it assumes, and it is never described as a measurement.

Next step → See every tool cost side by side

Price data last checked 2026-11-18 against official sources · how we verify

Price-change alerts

One short email when a provider changes a price or a promo ends. No spam, unsubscribe any time.