← All toolsDOSSIER · Build & code · FLAGSHIP · VERIFIED 2026-06-07

Cursor81FlagshipBenchmark pendingAI-native code editor and agentic refactor workspaceVerified 2026-06-07

Dossier · Build & code

Cursor

AI-native code editor and agentic refactor workspace · last verified 2026-06-07

FlagshipEDITOR'S PICK

Build & code

Cursor

81/100

ROLEAI-native code editor and agentic refactor workspace

Editorial fit

Source quality

Citation honesty

Privacy controls

Value for money

Speed

FREEHobby

$20/MOPro

$60/MOPro+

FlagshipVerified 2026-06-07VP·METHOD

Sixteen scored dimensions, Tab/Agent/Chat surfaces, Pro · Pro+ · Ultra · Teams pricing, competitive lens vs Factory · Copilot · Windsurf, and desk rows waiting for your repo benches.

What is Cursor?

Cursor is a tool in the VerdictPal Build & code set: VS Code-compatible AI editor with Tab completion, Chat, Agent, MCPs, skills, hooks, cloud agents, Bugbot, Privacy Mode, and team/enterprise governance controls. It scores 81 out of 100 on editorial fit.

Role: AI-native code editor and agentic refactor workspaceCategory: Build & codeEditorial · hands-on

Cursor at a glance, with the date each field was checked.
Field	Value
Editorial fit	81 out of 100
Dossier status	Flagship
Set	Build & code
Role in a workflow	AI-native code editor and agentic refactor workspace
Pricing	Hobby Free · Pro $20/mo · Pro+ $60/mo · Ultra $200/mo · Teams $40/mo · Enterprise Custom
Training on your data	Not recorded as a plan-wide guarantee. Check the dossier's privacy notes and sources.
Last verified	2026-06-07

Editor's note

Cursor is the strongest tool when the story is “AI helps me produce a diff I can inspect.” The tool should punish any workflow where the AI becomes an unreviewed committer.

Public facts · vendor & repo

List prices and pay-as-you-go entry points we can cite without running our own bench. Each tile links to a source when possible.

$20/moProcursor.com/pricing ↗

$60/moPro+cursor.com/pricing ↗

$200/moUltracursor.com/pricing ↗

$40/user/moTeamscursor.com/pricing ↗

12 mo freeStudent Procursor.com/students ↗

Limits & product surface

Non-price vendor claims — multipliers, caps, and API scope. Detailed matrices live in subscription and SDK sections below.

Annual discount: ~20% when billed yearlycursor.com/pricing ↗
Privacy Mode: No training / no retention (when enabled)docs.cursor.com ↗
SOC 2: Type II (vendor attestation)cursor.com/security ↗
Public dev API: None — IDE subscription onlyVerdictPal desk ↗
Editor base: VS Code / Code-OSS forkProduct positioning ↗

How it works · agent loop

The public positioning for this product — the loop we score against on VerdictPal.

Scope the Agent

Name files and tests before you accept a plan — vague prompts become repo-wide diffs.

Review every hunk

Apply is not merge. Run typecheck and tests on your machine before you commit.

Set Privacy Mode

Enable for course, client, and thesis repos unless your syllabus allows provider processing.

Watch the meter

Pin frontier models when quality matters; use Auto when you are exploring — pools are not unlimited on Pro.

Who it fits, where it fails

Best for

Developers already living in VS Code who want Tab, Chat, and Agent in one editor
Repo-wide refactors where multi-file diffs must be reviewed before merge
Teams that need SAML/OIDC, centralized billing, usage analytics, team-wide Privacy Mode, and admin controls
Students eligible for one free Pro year who can still run tests and read generated code

Avoid if

You cannot enable Privacy Mode or get approval before sending sensitive code to AI providers
You need deterministic flat pricing with no usage-based overage exposure
You want a headless agent SDK rather than a human-in-the-loop editor
You tend to accept green-looking multi-file diffs without running tests

Strengths

Tab completions
inline edits
Chat
Agent
MCPs
skills
hooks
rules
cloud agents
Bugbot
individual usage pools
Teams SSO/admin/analytics/privacy mode
Enterprise SCIM/audit/access controls/service accounts/AI code tracking API

Weaknesses

Agent edits can break tests/security despite clean diffs
Usage pools and on-demand billing are hard to forecast
Privacy Mode off can expose sensitive code context
Model/provider routing creates data and cost differences
MCP/cloud agents add permission and execution boundaries
Students may substitute AI edits for understanding

How it could improve

Over-refusal is a calibration problem. The tool needs finer-grained policy categories so legitimate research tasks don't get swept up in blanket safety filters.

Search modes · 8 lenses

One input box, many retrieval postures. Filter by tier.

01Free

Tab

Inline completions as you type — limited on Hobby, unlimited on paid tiers

02Free

Inline edit

Cmd+K selection edits with diff preview

03Free

Chat

Ask questions with @file and @docs context

04Pro

Agent

Multi-file plans and applies — extended limits on Pro+

05Pro

Cloud agents

Long-running remote agent jobs on paid tiers

06Pro

Bugbot

PR-style AI review — usage-based billing per pricing page

07Pro

Rules & MCP

Project rules, skills, hooks, and MCP connectors

08Pro

Auto model

Unlimited on paid tiers; routes models without burning frontier pool as fast

Competitive lens

Where Cursor wins for cited research — and where a rival still belongs in the stack.

Factory (Droid) is agent-native delivery with CLI, IDE, cloud agents, and an SDK; Cursor is the polished VS Code fork most students already installed. Pick Factory when the SDK and ZDR team contract matter; pick Cursor when diff-first editing in a familiar IDE matters.

Cursor

VS Code extension compatibility and muscle memory
Tab completion quality in daily typing loops
Larger student community and coursework tutorials
Privacy Mode story on individual and Teams seats

Factory

First-class SDK and headless agent surfaces
Bring-your-own keys and Zero Data Retention on Teams
Positioned for software delivery automation, not just editing
Clearer split for orgs that want agents without a GUI fork

Scores and evidence

Metric lab · 16 dimensions

Click a tile for the editorial note. Color follows score: coral, yellow, mint.

Shape

Avg 76 · 55–90

81Editorial fit

Default AI IDE for repo-native work; not a literature search tool or a headless delivery agent.

By group

Fit80

Editorial fit81
Agent depth83
Cloud agents70
Model breadth84

Cost67

Value for money72
Credit clarity62
Hobby tier utility55
Student offer80

Trust75

Diff discipline74
Privacy Mode76
Teams admin79
Ecosystem lock-in71

Workflow81

Tab completion86
VS Code compatibility90
Rules & MCP78
Bugbot reviews68

Benchmark ledger

Public rows are vendor or third-party claims we logged with a date. Desk rows are reserved for VerdictPal self-run results.

Measure	Result	Source
Pro subscription	20 USD / mo2026-06-07	cursor.com/pricing
Pro+ subscription	60 USD / mo2026-06-07	cursor.com/pricing
Ultra subscription	200 USD / mo2026-06-07	cursor.com/pricing
Teams	40 USD / user / mo2026-06-07	cursor.com/pricing
Pro usage pool (vendor)Credit-based billing since mid-2025; frontier model use draws down the pool.	20 USD / mo included2026-06-07	cursor.com/pricing
Pro+ usage vs Pro	3 × pool2026-06-07	cursor.com/pricing
Ultra usage vs Pro	20 × pool2026-06-07	cursor.com/pricing
Annual billing	20 % off (vendor)2026-06-07	cursor.com/pricing

Editorial evidence · 0 entries

No editorial benchmark yet. The tool stays at status Flagship until evidence lands.

Benchmark coverage · 8 benchmarks

AA output speed ⚠AA time to first token ⚠Agent Diff Review (VerdictPal Lab)DeepSWE HumanEval ⚠SWE-bench Verified ⚠Terminal-Bench-Science Terminal-Bench

Benchmarks that evaluate or report on this tool. Scores and caveats live on the benchmark page.

Test scenarios · hands-on lab

How we exercised the product. Step through each run before you trust the scores.

Test run

Step 1 of 5

Clone a fixed open-source fixture repo with passing tests.

Research log · desk notes

What we learned while building this dossier — not vendor copy.

2026-05-30cursor.com/pricing ↗

Credit pools replaced request caps

Since mid-2025, paid tiers include a usage pool sized to the plan price; pinning frontier models draws it down. Auto mode stays unlimited on paid tiers per vendor copy — syllabus should mention both.

2026-05-30cursor.com/pricing · Enterprise page ↗

No public developer API

Cursor is a desktop agent surface, not an HTTP API product. Enterprise lists an AI code tracking API for analytics — do not confuse with Sonar-style search APIs on other cards.

2026-05-30cursor.com/security ↗

Teams enforces Privacy Mode org-wide

Individual Privacy Mode is necessary but not sufficient for departments — Teams adds org-wide controls and SSO. Pair with your institution's cloud policy before assigning graded repos.

2026-05-30cursor.com/students ↗

Student Pro is twelve months

Verified students get Pro free for 12 months — still enable Privacy Mode on sensitive repos; the offer does not change provider routing by itself.

What it costs, what it keeps

Subscription · checked 2026-06-07

Plan matrix

Individual plans from cursor.com/pricing May 2026. Teams is the business tier ($40/user/mo). Usage pools scale Pro → Pro+ (3×) → Ultra (20×); Auto mode is unlimited on paid tiers per vendor copy.

Capability	Free	Pro	Max
Tab completions	Limited	Unlimited	Unlimited
Agent / Composer	Limited	Extended	Highest limits
Frontier usage pool	—	$20/mo equiv.	$200/mo equiv. (Ultra)
Auto mode	—	Unlimited	Unlimited
Cloud agents	—	Yes	Yes + priority features
MCP · skills · hooks	—	Yes	Yes
Privacy Mode	Per user	Per user	Per user
Teams admin / SSO	—	Teams plan	Teams + Enterprise

cursor.com individual and team seats May 2026. Pro is the default desk pick; Pro+ is vendor-recommended for daily Agent users; Ultra for power users; Teams is the business tier (often called Business in comparisons). No public HTTP API for app builders.

Desk pick

Pro

$20/mo

Monthly · ~20% off annual

Unlimited Tab on paid tier
Extended Agent limits
Frontier model pool (~$20/mo usage included)
MCP, skills, hooks, cloud agents
Privacy Mode available per user

Pro+

$60/mo

Monthly · ~20% off annual

Everything in Pro
3× usage on OpenAI, Claude, Gemini (vendor)
Vendor-recommended for daily Agent users

Ultra

$200/mo

Monthly · ~20% off annual

20× Pro usage on frontier models (vendor)
Priority access to new features
For agent-first workflows all day

Enterprise is custom (pooled usage, SCIM, audit logs, AI code tracking API for admins — not a coding agent API). Students: 12 months Pro free at cursor.com/students. Pro+ sits between Pro and Ultra when you routinely exhaust the $20 pool.

Privacy deep-dive · checked 2026-06-07

Privacy Mode can be enabled by users or team admins. With it on, Cursor says code data is not stored by model providers or used for training, with ZDR controls; Cursor may still store some code data for extra features. With Privacy Mode off, codebase data, prompts, editor actions, and snippets may be used to improve/train features.

Under the hood

Vendors named on the product about page — useful for procurement and privacy reviews.

Editor shell: VS Code / Code-OSS fork
Models: OpenAI · Anthropic · Google (routed)
Extensions: VS Code marketplace (mostly compatible)
Auth: Cursor account · SSO on Teams
Compliance: SOC 2 Type II (vendor claim)
Open source: No public editor fork

Alternatives and context

Alternatives

GitHub CopilotWindsurfClineZedFactory

Head-to-head

Cursor vs Factory

Compare side-by-side → open compare

Workflow roles

How this tool fits into a composed research stack:

code authoringagentic refactor

Powered by

The model(s) that drive this tool. The tool is the wrapper; the model is the engine.

Deep panes · 9 lenses

Editorial lenses only. Subscription and API pricing live in their own sections above.

Lens 01

Tab completion

Tab is the fastest surface and the easiest to over-trust. It excels at boilerplate; it will not infer your assignment rubric.

Desk rule: Accept sparingly on graded algorithms
Paid tier: Unlimited Tab on Pro+

Log when Tab wrote a function you cannot explain orally
Pair Tab with tests — accepts are not proofs

Sources and provenance

Source ledger · 6 checked

Pricingchecked 2026-06-07
Models & pricing docschecked 2026-06-07
Securitychecked 2026-06-07
Data usechecked 2026-06-07
Studentschecked 2026-06-07
Teams pricing updatechecked 2026-06-07

Quality gate · Flagship

Evidence ready

Benchmark pending

Verdict history

2026-06-07Cursor remains the AI-native code editor flagship, but the tool now needs to emphasize three separations: editor workflow versus headless agents, individual usage pools versus team/enterprise governance, and Privacy Mode versus ordinary provider transit. Cursor is excellent for repo-local Tab, Chat, Agent, MCP, cloud agents, and Bugbot workflows, but only if diffs are reviewed and tests run.
2026-06-17Production-readiness audit pass: editorsNote and pricing/privacy checked dates confirmed. benchmarkReady remains false intentionally until agent-diff safety, usage-burn, and Privacy Mode runs land.

Related dossiers

Build & code

Factory

74/100

ROLEAgent-native delivery surface

Agent-native software development platform around Droid, with terminal, desktop, CLI, SDK, local/cloud background agents, missions, review workflows, and enterprise controls for delegated software delivery.

Watch outYou only want lightweight VS Code autocomplete

Editorial · hands-onFlagship

Multi-step refactors, migrations,Teams wiring Droid

ROLETaste-aware terminal coding agent and open-model harness

Terminal coding agent with continuous taste learning (taste-1), open-model deals, MCP/skills, and subscription credits from $1/mo — ships, tests, and refactors in your conventions.

Watch outYou need a GUI IDE with Tab completion as the primary surface — use Cursor instead

Editorial · hands-onFlagship

Developers who wantBuilders experimenting DeepSeek,

ROLEPrompt-to-app builder and MVP prototyping tool

AI app-builder that turns natural-language prompts into web apps, projects, code, deployed apps, Lovable Cloud backends, AI Gateway usage, domains, and collaborative team workspaces.

Watch outYou need production-grade security without developer or security review

Editorial · hands-onSolid

Founders, students, productNon-specialists who can

ROLEOpen-source terminal coding agent and model-agnostic harness

Open-source AI coding agent for terminal, desktop, and IDE — 75+ model providers, Plan/Build modes, LSP support, optional OpenCode Go ($10/mo) and Zen pay-as-you-go gateway.

Watch outYou need a polished GUI IDE with Tab-as-primary workflow — Cursor may fit better

Editorial · hands-onFlagship

Developers who wantTeams that need

ROLEAI app-development workspace and hosted coding environment

Browser-based AI development platform for building, editing, deploying, collaborating on, and publishing apps with Replit Agent, hosted workspaces, built-in database, authentication, hosting, monitoring, integrations, and usage credits.

Watch outYou need full local control, predictable AI-agent spend, or mature production SRE workflows

Editorial · hands-onSolid

Students solo buildersFast prototype builds

ROLEAI-native terminal and agentic development workspace

Agentic development environment born from a modern terminal, with Terminal and Agent modes, local/cloud coding agents, third-party CLI agent orchestration, Warp Drive workflows, code review, and enterprise privacy controls.

Watch outYou require a fully local traditional terminal with no account/cloud features

Editorial · hands-onSolid

Developers who wantTeams experimenting local

Questions this dossier answers

Every answer below is assembled from the dated fields on this page. Nothing is written separately for search.

Is Cursor worth using?

Cursor scores 81 out of 100 on editorial fit and carries Flagship dossier status. Its job in a research workflow is: AI-native code editor and agentic refactor workspace.

Who is Cursor best for?

Cursor earns its place when you need:

Developers already living in VS Code who want Tab, Chat, and Agent in one editor
Repo-wide refactors where multi-file diffs must be reviewed before merge
Teams that need SAML/OIDC, centralized billing, usage analytics, team-wide Privacy Mode, and admin controls
Students eligible for one free Pro year who can still run tests and read generated code

When should you not use Cursor?

Skip Cursor in these cases:

You cannot enable Privacy Mode or get approval before sending sensitive code to AI providers
You need deterministic flat pricing with no usage-based overage exposure
You want a headless agent SDK rather than a human-in-the-loop editor
You tend to accept green-looking multi-file diffs without running tests

What does Cursor cost?

Hobby Free · Pro $20/mo · Pro+ $60/mo · Ultra $200/mo · Teams $40/mo · Enterprise Custom. Pricing last checked 2026-06-07.

Does Cursor train on your data?

Not recorded as a plan-wide guarantee. Check the dossier's privacy notes and sources.. Privacy terms last checked 2026-06-07.

What goes wrong with Cursor?

The dossier publishes 6 failure modes for Cursor, and they stay published whether or not the vendor likes them:

Agent edits can break tests/security despite clean diffs
Usage pools and on-demand billing are hard to forecast
Privacy Mode off can expose sensitive code context
Model/provider routing creates data and cost differences
MCP/cloud agents add permission and execution boundaries
Students may substitute AI edits for understanding

What are the alternatives to Cursor?

The closest options to Cursor are GitHub Copilot, Windsurf, Cline, Zed, Factory. Each one that has a tool in the atlas is linked from this dossier, with a head-to-head comparison.

The Pack · Editorial newsletter

New tools in your inbox. Free.

One short email when a tool ships or changes status. No tracking, no third-party analytics. Unsubscribe in one click.

Info

Quality gate

Evidence ready
Benchmark pending
Pricing checked 2026-06-07
Privacy checked 2026-06-07

Evidence

Editorial · hands-on

Tested by VerdictPal desk · 2026-05-30

6 sources · 4 bench entries

Verdict history

2026-06-07Cursor remains the AI-native code editor flagship, but the tool now needs to emphasize three separations: editor workflow versus headless agents, individual usage pools versus team/enterprise governance, and Privacy Mode versus ordinary provider transit. Cursor is excellent for repo-local Tab, Chat, Agent, MCP, cloud agents, and Bugbot workflows, but only if diffs are reviewed and tests run.
2026-06-17Production-readiness audit pass: editorsNote and pricing/privacy checked dates confirmed. benchmarkReady remains false intentionally until agent-diff safety, usage-burn, and Privacy Mode runs land.