Clarifying our pricing

原始链接: https://cursor.com/en/blog/june-2025-pricing

Cursor acknowledges that recent pricing changes for individual Pro plans were poorly communicated, causing confusion and frustration. They are issuing refunds for unexpected charges incurred between June 16th and July 4th due to this miscommunication; contact [email protected] for a full refund. The new Pro plan includes unlimited use of Tab and models in "Auto" mode, plus $20 of monthly usage credit for other frontier models at API pricing, with the option to purchase more. Cursor shifted from request-based to API-based pricing to better reflect the varying costs of different requests. The initial announcement failed to clarify that "unlimited usage" was limited to the "Auto" feature, leading users to believe it applied to all models. Cursor recognizes this misstep and promises to improve communication regarding future pricing adjustments, including advance notice, clearer documentation, and enhanced customer support. They have already updated the pricing page and dashboard for greater visibility of usage limits.

A Hacker News thread discusses Cursor's pricing changes and their business viability in the evolving AI landscape. Several users express concerns about Cursor's future, pointing to its reliance on LLMs they don't own and increasing competition from in-house coding solutions from providers like Claude. Some users believe Cursor over-raised capital and question the value of their wrapper around existing LLMs, especially given the rise of more cost-effective and powerful solutions like Claude Code and open-source alternatives such as Roo Code and AMP Code. Others argue that Cursor's value lies in its fast autocomplete powered by SuperMaven and integrated user experience, while some suggest newer models are getting larger and more resource heavy through reasoning, meaning costs per request are going up instead of down. The discussion also touches on the broader implications of rising compute costs and the potential impact on AI-driven startups that rely on models becoming cheaper.

原文

Our recent pricing changes for individual plans were not communicated clearly, and we take full responsibility. We work hard to build tools you can trust, and these changes hurt that trust.

Before we break down how the new Pro pricing works, we want to make sure you know that we will refund any unexpected charges you may have incurred for usage over the past 3 weeks.

Timeline:

June 16: Original post released, changes made to Pro plan and Ultra introduced, Teams plans unchanged.
June 30: We updated the original post and our pricing page to improve their clarity.
July 4: We want to apologize and provide more details in this post.

How the new Cursor Pro works

The new Cursor Pro plan gives you:

Unlimited usage of Tab and models in Auto
$20 of frontier model usage per month at API pricing
An option to purchase more frontier model usage at cost

We previously described our included model usage as "rate limits", which wasn't an intuitive way to describe the limit. It is a usage credit pool for the month. After you've exhausted your credit, we still give a short grace period beyond the credit pool.

If you exceed your included usage, and do not want to use models in Auto, you can enable a spend limit. This allows you to pay for additional usage at cost.

Why did we change the plan?

Cursor uses a combination of our custom models, as well as models from providers like OpenAI, Anthropic, Google, and xAI. For external models, we previously charged based on the number of requests made. There was a limit of 500 requests per month, with Sonnet models costing two requests.

New models can spend more tokens per request on longer-horizon tasks. Though most users' costs have stayed fairly constant, the hardest requests cost an order of magnitude more than simple ones. API-based pricing is the best way to reflect that.

We moved from request-based pricing to $20 of included usage on Pro. We also started to offer unlimited usage of models when Auto is selected. Auto automatically routes to different frontier models based on capacity.

We were not clear that "unlimited usage" was only for Auto and not all other models, which have at least $20 of included usage. Based on median token usage, the Pro plan currently covers about 225 Sonnet 4 requests, 550 Gemini requests, or 650 GPT 4.1 requests, and the vast majority of Pro users do not run out of their included usage.

We missed the mark

We recognize that we didn’t handle this pricing rollout well and we’re sorry. Our communication was not clear enough and came as a surprise to many of you. We’re improving how we communicate future pricing changes.

We are offering usage refunds for those with unexpected usage between June 16 and July 4. Please contact [email protected] if you had a surprise usage bill and the team will work as fast as possible to get you a full refund.

What we’re shipping next

We’re committed to doing a better job with pricing updates going forward. This means advance notice, clearer documentation, and ensuring our team is prepared to support you through transitions.

We’ve improved our pricing page, documentation, and added additional visibility on the Cursor Dashboard to see when you’re approaching usage limits. Expect more updates in the coming weeks.