Cursor logo

Cursor Pricing

All plans and pricing as of 2026-07-04

Free tier available6 plansGreat value score (8/10)

Hobby (Free)

$0
  • 2,000 completions
  • 50 premium requests
  • Composer 2 limited access
Start Free
Most Popular

Pro

$20/month
  • Unlimited completions
  • Extended agent limits
  • Composer 2 Standard at $0.50/$2.50 per 1M tokens
  • All frontier models
Get Pro

Pro+

$60/month
  • 3x Pro usage on OpenAI, Claude, Gemini models
  • Composer 2 Fast at $1.50/$7.50 per 1M tokens
  • Multi-workspace
Get Pro+

Ultra

$200/month
  • 20x Pro usage on all frontier models
  • Highest priority queue
  • Background Agents at scale
Get Ultra

Teams -- Standard seat (new pricing, renewals 2026-07-01)

$32-40/user/month ($32 annual / $40 monthly)
  • Now includes MORE total usage at the same price (leverages Composer 2.5 cost efficiency)
  • Two separate usage pools: (1) first-party Composer/Auto, (2) third-party API models
  • Shared chats, centralized billing, analytics, SAML SSO, privacy mode, Agent Command Center
Get Teams -- Standard seat (new pricing, renewals 2026-07-01)

Teams -- Premium seat (new, 2026-07-01)

$96-120/user/month ($96 annual / $120 monthly)
  • 5x the included usage of the Standard seat at only 3x the cost
  • Same two-pool structure (first-party vs third-party API)
  • For heavy agentic teams that were previously overflowing the Standard seat
Get Teams -- Premium seat (new, 2026-07-01)

Is Cursor Worth the Price?

A

Value Score: 8/10

Overall Score: 8.3/10 · Developers who want the deepest AI integration possible and who are ready to work with agents rather than just autocomplete. Cursor 3's multi-workspace + cross-platform agent story is designed for people who are already living in the Cursor app daily, not dabblers.

Cursor 3 + Composer 2 (April 2026) is the release where Cursor stops being an AI-augmented editor and becomes an agent-first dev environment. Composer 2's speed at $0.50/$2.50 per 1M tokens is the real story -- it makes tight iteration loops feel instant in a way frontier models can't, even if the quality is slightly below Opus 4.6. The multi-workspace + cross-platform handoff design assumes you want your agents everywhere; if you don't, it's overkill. The honest read: for developers willing to go all-in on agentic workflows, Cursor 3 is the 2026 default. For everyone else, Claude Code + Copilot is a simpler, less committing stack.

The Tier List Tuesday

Weekly newsletter: tier movers, new entrants, and the VS of the week. Built from our daily AI-tool sweeps. No spam, unsubscribe anytime.

How Cursor Pricing Compares

ToolFree TierStarting PriceValue ScoreOverall
Cursor(this tool)Yes$08/108.3
GitHub CopilotYes$08/108.3
Codex (OpenAI)Yes$08/108.3
Google AntigravityYes$06/108.0
Roblox AssistantYes$09/108.0
Claude CodeNo$20/month7/107.8