Agent productExpanded profile
Cursor
Coding agent available through desktop and CLI interfaces.
Official link: cursor.com (opens in a new tab)Sources checked November 13, 2026
Source-linked research and editorial guidance. Capabilities are not independently tested unless a profile explicitly includes a reproducible test record. Sources checked November 13, 2026.
Evidence status
- Description
- Source summary — not runtime-tested
- Testing
- Not runtime-tested No runtime test, payment test or security audit was performed for this release. A source shows that something is documented, not that it was independently tested.
- Identity
- Candidate record; identity not fully resolved
- Evidence note
- Capabilities are vendor-described; task success, permissions and autonomous execution were not independently tested.
What it does
Capabilities
Vendor-documentedSource: cursor.com (opens in a new tab)
Pricing
Snapshot from the cited source. Recheck current terms and total workflow cost.
Pricing snapshot
Vendor-documentedThe inspected page lists Free, Plus at $8 per month, Premium at $30, Super at $100 and Ultra at $200. Plus is described as limited to Grok models. Third-party model usage is billed separately.
Plan prices are not a cost per completed task.
Source: cursor.com/pricing (opens in a new tab) (observed November 13, 2026; currency USD)
How to evaluate it
Editorial guidance and inference from the atlas — not a claim by the vendor and not a test result.
Compare the models and execution surfaces included in your chosen plan. A subscription price alone does not describe total inference spend.
Evidence gaps
What the inspected sources did not establish. Unknown does not mean absent or unsupported.
- Documented interfaces (MCP or other): Not established by the inspected sources
- People, organization and funding: Not established by the inspected sources
- Repository: Not established by the inspected sources
Verification backlog High priority
- Approval policy
- Runtime compatibility
- Independent task results
- Recheck volatile prices and versions at publication
Related records to investigate
Linked by topic; not a ranking or a verified compatibility claim.
Category: Coding agents
Listed as a research starting point (not a tested recommendation) in: How to evaluate an agent for fixing a GitHub issue
Cognition
Operates Devin for planning, writing, testing and shipping code.Replit
App-building platform with an Agent product and deployment infrastructure.Factory
Enterprise coding agents spanning planning, implementation, review and shipping.CodeRabbit
Automated pull-request review and agentic change management.Augment Code
Cosmos platform coordinates software delivery from specification through verification.Sources
Spotted something out of date? Suggest a correction