What Cursor costs. And Claude Opus 5.5 for 20 to 25% less past your limits.

Hobby is free; Pro is $20 a month, Pro Plus $60 and Ultra $200. On those three, on-demand usage past the plan bills Claude Opus 5.5 at $4 / $20 per million input and output tokens; on your own APIsRouter key, $3 / $16.

Example: a month of Agent requests on Claude Opus 5.5 past the included usage

Example request: 25,000 input and 2,000 output tokens, no cached input; real requests vary. Cursor estimates that people who run the Agent every day use $60 to $100 a month in total.

Cursor on-demandAPIsRouter key
Light use, 60 requests$8.40$6.42
Every workday, 400 requests$56$42.80
Agents all day, 2,000 requests$280$214

Every Cursor plan

Paid plans include a set amount of model usage each month. Past it, you turn on on-demand usage, billed at the model rates below, or upgrade.

Plans from cursor.com/pricing and cursor.com/docs/models-and-pricing, September 2026, in US dollars before tax.
PlanPriceWhat it includesSuits
Hobby$0Limited Agent requests and Composer, no credit cardTrying Cursor
Pro$20 a monthUnlimited Tab, extended Agent limits, frontier models, cloud agentsAgent use now and then
Pro Plus$60 a monthEverything in Pro, with 3 times Pro's Agent limitsRunning the Agent every day, as Cursor recommends
Ultra$200 a monthEverything in Pro, with 20 times Pro's Agent limits and priority access to new featuresHeavy Agent use, as Cursor recommends
Teams$40 or $120 per user a monthTeam billing, SSO and Bugbot. Premium seats get 5 times the Agent limits of StandardTeams shipping together

Cursor's per-token rates and an APIsRouter key

On Pro, Pro Plus and Ultra, on-demand usage past the included usage bills third-party models at these rates. Requests on your own API key skip the included usage and bill per token with that key. Shown are the models APIsRouter sells below both Cursor's rate and the maker's own price, on input, cached input and output. Cursor rates from cursor.com/docs/models-and-pricing, September 2026, per million tokens.

ModelInput: Cursor / APIsRouterCached input: Cursor / APIsRouterOutput: Cursor / APIsRouter
Claude Fable 5.1 (claude-fable-5-1)$10 / $8$0.25 / $0.20$50 / $40
Claude Fable 5 (claude-fable-5)$10 / $8$1 / $0.80$50 / $40
Claude Opus 5.5 (claude-opus-5-5)$4 / $3$0.20 / $0.15$20 / $16
Claude Opus 5 (claude-opus-5)$5 / $4$0.50 / $0.40$25 / $20
Claude Opus 4.8 (claude-opus-4-8)$5 / $4$0.50 / $0.40$25 / $20
Claude Opus 4.7 (claude-opus-4-7)$5 / $4$0.50 / $0.40$25 / $20
Claude Opus 4.6 (claude-opus-4-6)$5 / $4$0.50 / $0.40$25 / $20
Claude Sonnet 5.5 (claude-sonnet-5-5)$2 / $1.50$0.20 / $0.15$10 / $8
Claude Sonnet 5 (claude-sonnet-5)$2 / $1.50$0.20 / $0.15$10 / $8
Claude Sonnet 4.6 (claude-sonnet-4-6)$3 / $2.40$0.30 / $0.24$15 / $12
Claude Haiku 4.5 (claude-haiku-4-5-20251001)$1 / $0.80$0.10 / $0.08$5 / $4
GPT-5.6 Sol (gpt-5.6-sol)$4 / $3$0.40 / $0.30$20 / $16
GPT-5.6 Terra (gpt-5.6-terra)$2 / $1.50$0.20 / $0.15$12 / $9.50
GPT-5.6 Luna (gpt-5.6-luna)$0.20 / $0.15$0.02 / $0.015$1.20 / $0.95
GPT-5.5 (gpt-5.5)$5 / $4$0.50 / $0.40$30 / $24
Gemini 3.8 Flash (gemini-3.8-flash)$0.75 / $0.60$0.075 / $0.06$3.50 / $3
Gemini 3.7 Flash (gemini-3.7-flash)$0.75 / $0.60$0.075 / $0.06$3.50 / $3
Gemini 3.6 Flash (gemini-3.6-flash)$1.50 / $0.60$0.15 / $0.06$7.50 / $3
Gemini 3.5 Flash (gemini-3.5-flash)$1.50 / $1.20$0.15 / $0.12$9 / $7
Gemini 2.5 Flash (gemini-2.5-flash)$0.30 / $0.24$0.03 / $0.024$2.50 / $2
Kimi K3 (kimi-k3)$3 / $2.50$0.30 / $0.25$15 / $13.50
Kimi K2.7 Code (kimi-k2.7-code)$0.95 / $0.85$0.19 / $0.15$4 / $3.50
GLM-5.3 (glm-5.3)$1.40 / $1.20$0.26 / $0.20$4.40 / $3.50
GLM-5.2 (glm-5.2)$1.40 / $1.20$0.26 / $0.20$4.40 / $3.50
GLM-5.3-Flash (glm-5.3-flash)$0.15 / $0.13$0.029 / $0.025$0.50 / $0.45

Models Cursor's rate list does not include

Add these to Cursor through APIsRouter as custom models: GPT-6 Astra, GPT-6 Sol, DeepSeek V4 Pro and Qwen3.8-Max, each below its maker's price. per million input / output tokens

ModelMaker's priceAPIsRouter
GPT-6 Astra (gpt-6-astra)OpenAI: $10 / $50$8 / $40
GPT-6 Sol (gpt-6-sol)OpenAI: $2 / $10$1.50 / $8
DeepSeek V4 Pro (deepseek-v4-pro)DeepSeek: $0.66 / $1.98 (Off-peak prices on both sides)$0.55 / $1.50
Qwen3.8-Max (qwen3.8-max)Alibaba Cloud: $2 / $6 (Alibaba Cloud's international price)$1.40 / $4

Add APIsRouter to Cursor

  • Get a key: Start free on the $2 trial credit, or top up from $10 below. A top-up key arrives by email.
  • Paste the key and the base URL: In Cursor Settings > Models, paste the key under OpenAI API Key, turn on Override OpenAI Base URL and enter https://api.apisrouter.com/v1.
  • Add models by ID: Add each model by its exact ID, such as claude-opus-5-5, then pick it in chat or the Agent. Tab completion stays on Cursor's own models. While the override is on, Cursor also sends its built-in OpenAI models to APIsRouter, so turn it off to use them on your plan.
OpenAI API Key             sk-your-apisrouter-key
Override OpenAI Base URL   https://api.apisrouter.com/v1
Custom model               claude-opus-5-5

Two ways to pay

Both fill one balance, spent at the APIsRouter prices above.

FAQ

How much does Cursor cost?

Hobby is free. Pro costs $20 a month, Pro Plus $60 and Ultra $200, before tax. Teams cost $40 per user a month on Standard seats and $120 on Premium; Enterprise is priced on request.

Is Cursor free?

Yes, on Hobby: no credit card, limited Agent requests and access to Composer. Pro and above add unlimited Tab, extended Agent limits, frontier models and cloud agents.

Pro, Pro Plus or Ultra?

Pro Plus has 3 times Pro's Agent limits and Ultra 20 times; Cursor recommends Pro Plus for people who run the Agent every day and Ultra for heavy Agent use. All three include unlimited Tab and two usage pools, one for Cursor's own models and one for third-party models such as Claude Opus 5.5.

What happens when I hit the usage limit?

You can turn on on-demand usage, billed monthly at the same per-token rates, or upgrade. On Pro, Pro Plus and Ultra you can also add your own API key: those requests skip the included usage, and on an APIsRouter key Claude Opus 5.5 costs $3 / $16 per million input and output tokens instead of $4 / $20.

Can I use my own API key in Cursor?

Yes. In Settings > Models you can add keys for OpenAI, Anthropic, Google, Azure OpenAI and AWS Bedrock, or point Override OpenAI Base URL at an OpenAI-compatible provider such as APIsRouter. Own keys work for chat models, and custom models added through the override also work in the Agent; Tab completion stays on Cursor's models. On Teams and Enterprise, Cursor adds $0.25 per million tokens to own-key requests.

Which models can I use in Cursor through APIsRouter?

Any text model in the APIsRouter catalog, added as a custom model: Claude Opus 5.5, Claude Fable 5.1, GPT-5.6 Terra and Kimi K3 cost less than Cursor's rate, and GPT-6 Astra, GPT-6 Sol, DeepSeek V4 Pro and Qwen3.8-Max are not on Cursor's rate list.