Stack L8
Cursor
AI-native IDE at $1B ARR; used across >50% of Fortune 500, now on in-house Composer model
HQ US
Fit Score
avg of 4 segmentsRecent Business Signals
Strategic Position
What they do best
Converting concentrated enterprise developer adoption into revenue — >50% of Fortune 500 engineers (signal 614e1292) underpinning ~$1B ARR (signal f5df3466).
Their bet
Owning the model layer: Composer (signal 9c590a6b) shifts Cursor from reseller of OpenAI/Anthropic capacity to a vertically integrated coding-agent vendor controlling cost and latency.
Top risk
Composer underperforms frontier third-party models in the next 6 months, so enterprises keep routing to Anthropic/OpenAI and Cursor's in-house margin/latency thesis fails to convert into ARR retention.
Compared to peers
Direct competitor
Cognition
Cognition's Devin sells autonomous agent execution; Cursor sells a human-in-the-loop IDE with >50% Fortune 500 seat penetration (signal 614e1292).
Substitute
Cognition
Agent-orchestration governance (Devin) can displace per-seat IDE spend by moving work off interactive editors entirely; Cursor counters with its 2.0 multi-agent interface (signal 9c590a6b).
Why someone would join
- 1.Revenue went from $500M annualized (Series C, 2025-06-05, signal 71a4d6a2) to ~$1B with a $29.3B Series D valuation by Nov 2025 (signal f5df3466) — sub-2-year doubling against real enterprise demand.
- 2.Hiring is concentrated in enterprise GTM and infra: 3 enterprise roles (AE, Forward Deployed Engineer, Enterprise PM, Feb–Apr 2026) plus a staff Inference Infrastructure Engineer (2026-05-06) to operate the in-house Composer model.
Recent Hiring (60 days)
- enterprise sales1
- developer tools1
- SaaS1
- enterprise product1
- admin controls1
Reverse-Hype Watch
- !Composer 'first in-house frontier coding model' that 'completes most agent turns in under 30 seconds' is vendor-stated (2025-10-29 Cursor 2.0) with no independent benchmark in signals — unverified differentiation claim.
- !Capacity/quality claim unverified: Cursor 2.0 'Composer, its first in-house frontier coding model, which the company said completes most agent turns in under 30 seconds' is vendor-stated with no independent benchmark in the signals.
- !Cursor 2.0 raw_text: Composer 'its first in-house frontier coding model, which the company said completes most agent turns in under 30 seconds' — performance/'frontier' claim is vendor-stated with no independent benchmark in the signals.
- !Capacity/performance claim unbacked by independent signal: Cursor 2.0 'Composer, its first in-house frontier coding model, which the company said completes most agent turns in under 30 seconds' — vendor-stated, no independent benchmark in signals.