Cursor
AI code editor and coding-agent platform with repository-aware workflows and multiple model providers.
Core record
Status: Public
Access: Desktop / CLI / Cloud agents
API status: Product integrations / SDK-dependent
Openness: Closed
Modalities: Code and text
Best suited for
Repository-aware coding agents, feature development, debugging, review and developer automation.
Limitations
Available models, routing behavior, pricing and context windows change over time.
API / auth
Interface: Developer editor / agents / product integrations
Base URL: Pending
Authentication: Cursor account / workspace authentication
Compatibility: Uses first-party Cursor models plus third-party model providers.
Rate limits: Usage pools, subscription plan and selected model determine limits.
Deployment
Models / versions
| Model | Context | Max output | Modalities | License | Status |
|---|---|---|---|---|---|
| Composer 2.5 composer-2.5 | Pending | Pending | code, text | Pending | current |
| Cursor Grok 4.6 | Pending | Pending | code, text | Pending | current |
Pricing snapshots
| SKU / model | Unit | Input | Output | Flat | Notes |
|---|---|---|---|---|---|
| Cursor Pro | USD / month | — | — | 20 | |
| Cursor Pro Plus | USD / month | — | — | 60 | |
| Cursor Ultra | USD / month | — | — | 200 | |
| Composer 2.5 | USD per 1M tokens | 0.5 | 2.5 | — | Cursor Models pool token rate shown on current docs. |
Integrations
Regions
Cursor documents optional regional data residency for eligible models/functions with pricing uplift.
Governance / privacy
Third-party models may be hosted by the provider, a trusted partner or Cursor.
Regional data residency pricing/eligibility is model-specific.
Benchmarks
Pending methodology-aware benchmark audit unless separately documented. Vendor and independent benchmark results are kept distinct.