Cursor just crossed $1 billion in annualised revenue. Over a million developers are paying for it. And a significant chunk of those developers are frustrated by a billing system they don't fully understand.
In June 2025, Cursor replaced its simple "500 fast requests per month" model with a credit-based system where costs vary depending on which AI models you use and how complex your requests are. For heavy users, what felt like an unlimited plan suddenly wasn't.
This guide breaks down every Cursor plan, explains exactly how the credit system works, and helps you figure out which tier is actually right for how you code, without any surprise charges at the end of the month.
Cursor Plans at a Glance
| Plan | Price | Best For |
|---|---|---|
| Hobby | Free | Trying Cursor before committing |
| Pro | $20/month | Solo developers, daily coding use |
| Pro+ | $60/month | Heavy AI users hitting Pro limits |
| Ultra | $200/month | Full-time AI-native development |
| Teams | $40/user/month | Engineering teams needing shared billing |
| Enterprise | Custom | Large orgs needing compliance and controls |
Annual billing saves 20% across all paid tiers.
The Hobby Plan (Free)
Cursor's free tier is genuinely useful for evaluation, but it's not designed for long-term use. You get limited Agent requests and limited Tab completions, but enough to experience Cursor's core features before deciding whether to pay.
No credit card is required to start, which is rare for a tool at this level. There's also a 7-day free Pro trial available if you want to test the full paid experience before committing.
Who it's for: Developers who want to try Cursor before switching from VS Code or GitHub Copilot. Don't expect to use it as your main editor on the free plan — you'll hit limits mid-feature.
The Pro Plan ($20/month)
Pro is Cursor's most popular plan and the entry point for serious use. At $20/month, you get:
- Unlimited Tab completions
- Extended limits on Agent requests
- Cloud Agents (background tasks that run while you work on something else)
- Maximum context windows
- Access to all frontier models (Claude, GPT-4, Gemini)
The shift to credit-based billing is where it gets nuanced. Your $20 subscription includes a $20 credit pool. Simple Tab completions and Auto mode usage don't touch your credits meaningfully. But if you manually select premium models like Claude Sonnet or GPT-4o for complex multi-file tasks, those eat into your pool faster.
Under the old system, Pro users got roughly 500 fast requests. Under the credit system, the equivalent is closer to 225 premium requests before overages kick in — though the number varies significantly based on task complexity and model choice.
The practical reality: If you mostly use Cursor for autocomplete and targeted prompts on familiar code, Pro is excellent value. If you're running multi-file refactors and architecture reviews with premium models every day, you'll hit limits and face overage charges.
Who it's for: Individual developers spending 2-4 hours daily in Cursor, using a mix of auto-suggestions and occasional Agent tasks.
The Pro+ Plan ($60/month)
Pro+ is Cursor's recommended tier and the one that makes the most financial sense for heavy users.
You get everything in Pro, plus 3x the usage credits. That means roughly three times as many premium model requests before overages apply. There are no new features compared to Pro; the entire upgrade is headroom.
The maths are straightforward: if you're regularly hitting Pro limits and paying $20-40 in monthly overages, Pro+ saves you money. If you're occasionally hitting limits, Pro is still the better value.
Who it's for: Developers spending 4+ hours daily in Cursor, regularly running multi-file edits, Agent tasks, and complex refactors with premium models.
The Ultra Plan ($200/month)
Ultra gives you 20x the usage credits of Pro, plus priority access to new features.
This is built for developers who live inside Cursor all day, every day. Running background agents continuously, working on large codebases with extensive context, and relying on frontier models for every interaction. At $200/month, it's infrastructure spend rather than a productivity tool subscription.
Who it's for: Full-time AI-native developers, vibe coders building entire products in Cursor, and anyone whose Pro+ overages regularly exceed $140/month.
The Teams Plan ($40/user/month)
Teams is where Cursor moves into proper business tooling. At $40/user/month, each user gets Pro-equivalent AI access, plus a layer of organisational features:
- Shared chats, commands, and rules across the team
- Centralised billing (one invoice, not individual subscriptions)
- Usage analytics and reporting
- Org-wide privacy mode controls
- Role-based access control
- SAML/OIDC SSO for enterprise identity management
- AWS Bedrock integration
The $20 included usage per user per month is worth noting, it's less than individual Pro plans, which matters for teams where not everyone codes intensively every day.
Who it's for: Engineering teams of 3+ developers who need shared context, centralised billing, and visibility into AI tool usage across the organisation.
Enterprise (Custom Pricing)
Enterprise adds the compliance and control layer that large organisations require:
- Pooled usage across the org (rather than per-user limits)
- Invoice/PO billing
- SCIM seat management
- AI code tracking API and audit logs
- Granular admin and model controls
- Priority support and account management
- Conversation Insights and Cursor Blame features
Contact Cursor's sales team for pricing. It's negotiated based on seat count and requirements.
Who it's for: Companies with SOC 2 compliance requirements, legal review of AI tool usage, or security teams who need visibility into what code is being shared with AI models.
The Bugbot Add-On
Cursor also offers Bugbot as a separate add-on, an AI code review tool that integrates with GitHub to automatically catch bugs in pull requests.
- Free: Limited reviews per month, GitHub integration, Cursor Ask access
- Pro: $40/user/month — unlimited reviews on up to 200 PRs/month, Bugbot Rules
- Teams: $40/user/month — unlimited reviews on all PRs, analytics dashboard
- Enterprise: Custom
Bugbot is separate from your main Cursor subscription and priced on top of your existing plan.
How the Credit System Actually Works
This is the part that has caused the most confusion — and the most Reddit threads.
Every paid Cursor plan includes a credit pool equal to your subscription cost. When you use AI features, credits are consumed based on the model you're using and the complexity of the request.
What doesn't use credits significantly:
- Tab completions (autocomplete)
- Auto mode (Cursor chooses the model automatically)
What consumes credits quickly:
- Manually selecting Claude Sonnet, GPT-4o, or Gemini Pro
- Large context windows (pointing Cursor at many files)
- Complex multi-file Agent tasks
- Background Agent runs
When your credit pool is exhausted, you can either pay for additional usage at API rates or switch to Auto mode for the rest of the month. Auto mode remains functional but uses lower-cost models.
The key insight: if you stick to Auto mode and Tab completions, Pro is effectively unlimited for most use cases. The credit system only bites when you're manually reaching for the most expensive frontier models on heavy tasks.
Cursor vs GitHub Copilot: The Pricing Reality
The obvious comparison is GitHub Copilot at $10/month individual, half the price of Cursor Pro.
The difference isn't arbitrary. Cursor is a full IDE replacement (built on VS Code) with deep codebase context, multi-file editing via Composer, and background Agents that can run tasks autonomously. Copilot is a plugin that adds AI assistance to your existing setup.
If you value your existing VS Code configuration and want AI assistance without switching editors, Copilot is the smarter value. If you want the AI woven into every part of your coding workflow (and you're willing to do the migration) Cursor justifies the premium for most full-time developers.
Windsurf is the closest direct competitor, offering a similar AI-native IDE at $15/month individual and $30/user for teams. It's become the default destination for developers frustrated with Cursor's pricing changes.
Which Plan Should You Choose?
Start with Hobby if you've never used Cursor and want to evaluate it before switching editors.
Go Pro if you code daily and want AI assistance throughout your workflow. Most developers find Pro sufficient unless they're running heavy Agent tasks.
Upgrade to Pro+ if your Pro overages regularly exceed $20-40/month. The maths favour Pro+ once you're hitting limits consistently.
Consider Ultra only if your Pro+ overages are substantial and you're treating Cursor as infrastructure rather than a subscription.
Choose Teams the moment you need centralised billing or shared context across a team of developers.
Talk to Enterprise if you have compliance requirements, security reviews, or need audit trails of AI tool usage.
The Student Discount
Cursor offers free access for students. Sign up with a school email address and verify on their student page. If you already have a paid plan, Cursor will refund the remaining balance automatically.
Is Cursor Worth It in 2026?
For developers who use it as their primary editor, yes, but with caveats.
The codebase-aware context and multi-file editing are genuinely difficult to replicate with other tools. Developers who've switched consistently report meaningful time savings on complex refactors and feature implementation.
The pricing controversy is real but manageable once you understand it. Staying on Auto mode and being selective about when you reach for premium models keeps costs predictable. The developers getting surprised bills are typically those running heavy Claude or GPT-4o sessions without monitoring their usage dashboard.
The practical advice: start on Pro, watch your usage dashboard for the first month, and upgrade only if you're consistently hitting limits. Most developers find $20/month is exactly right.
Pricing verified against Cursor's official pricing page, February 2026. Prices may change - check cursor.com/pricing for the latest.