Skip to content
Slow Variables
Ask

Deployment & application

Application-company revenue run-rates

Annualised revenue run-rates, ARR and annual revenue for leading AI application companies (Harvey, Cursor, Glean, Sierra, Decagon, Abridge, Hebbia, Perplexity, Replit, Lovable, Legora, Cognition, OpenEvidence, Clay), hand-entered from company statements, press reports and Sacra's estimates, each checked against its page.

Reason through this

The reading

Several application companies report run-rates in the hundreds of millions within two or three years of launch, but every figure is a company statement, a press leak or an analyst's estimate, and definitions differ (run-rate, ARR, contracted ARR, bookings). The direction reads Anysphere (Cursor) through one compiler, so the other companies stay as evidence rather than votes.

Every readingeach dot links to its record
a readingdisputedthe rule needs 3 readings; 2 so far
reported

2 readings drawn from dealroom.anysphere.revenue_run_rate_usd.pt, one of 36 series listed under How it is measured; 41 observations in the evidence.

Source: Dealroom news. Dealroom.

The numbers
as ofreadingvaluerecord
2026-06-05anysphere (disputed)$4.0Breported
2026-04-30anysphere (disputed)$3.0Breported

Why it matters

Whether the application layer keeps a share of the surplus depends on its revenue growing as model prices fall. Until now the tracker had no revenue for any application company.

What cuts against it

No figure is audited; run-rates annualise one month; companies choose when to disclose, so the series over-represents success; Sacra's figures are estimates.

How it is measured
  • series dealroom.anysphere.revenue_run_rate_usd.pt
  • series cnbc.harvey.arr_usd.pt
  • series pymnts.harvey.revenue_run_rate_usd.pt
  • series lawnext.harvey.arr_usd.pt
  • series techstartups.harvey.arr_usd.pt
  • series techcrunch.anysphere.revenue_run_rate_usd.pt
  • series forbes.anysphere.revenue_run_rate_usd.pt
  • series glean_press.glean.arr_usd.pt
  • series techcrunch.glean.arr_usd.pt
  • series techcrunch.sierra.arr_usd.pt
  • series sierra_blog.sierra.arr_usd.pt
  • series newcomer.sierra.revenue_run_rate_usd.pt
  • series sacra_free.decagon.revenue_run_rate_usd.pt
  • series newcomer.decagon.revenue_run_rate_usd.pt
  • series techcrunch.abridge.arr_usd.pt
  • series sacra_free.abridge.arr_usd.pt
  • series techcrunch.hebbia.arr_usd.pt
  • series sacra_free.hebbia.arr_usd.pt
  • series yahoo_finance.perplexity.arr_usd.pt
  • series yahoo_finance.perplexity.revenue_run_rate_usd.pt
  • series wallst_247.perplexity.revenue_run_rate_usd.pt
  • series techcrunch.replit.revenue_run_rate_usd.pt
  • series cnbc.replit.arr_usd.pt
  • series sacra_free.replit.revenue_run_rate_usd.pt
  • series lovable_blog.lovable.arr_usd.pt
  • series techcrunch.lovable.arr_usd.pt
  • series techcrunch.lovable.revenue_run_rate_usd.pt
  • series elevenlabs_blog.elevenlabs.arr_usd.pt
  • series techcrunch.elevenlabs.arr_usd.pt
  • series legora_newsroom.legora.arr_usd.pt
  • series runtimewire.legora.arr_usd.pt
  • series techcrunch.cognition.revenue_run_rate_usd.pt
  • series cnbc.openevidence.revenue_run_rate_usd.pt
  • series sacra_free.openevidence.revenue_run_rate_usd.pt
  • series clay_blog.clay.arr_usd.pt
  • series sacra_free.clay.arr_usd.pt
  • source Clay blog (GTM with Clay) · default tier 5 · Clay; short quotation
  • source CNBC · default tier 5 · CNBC; short quotation
  • source Dealroom news · default tier 5 · Dealroom; short quotation
  • source ElevenLabs blog · default tier 5 · ElevenLabs; short quotation
  • source Forbes · default tier 5 · Forbes; short quotation
  • source Glean press releases · default tier 5 · Glean; short quotation
  • source LawSites (LawNext) · default tier 5 · LawSites; short quotation
  • source Legora newsroom · default tier 5 · Legora; short quotation
  • source Lovable blog · default tier 5 · Lovable; short quotation
  • source Newcomer (free preview) · default tier 5 · Newcomer; short quotation of the free preview
  • source PYMNTS · default tier 5 · PYMNTS; short quotation
  • source RuntimeWire · default tier 5 · RuntimeWire; short quotation
  • source Sacra company pages (free summaries) · default tier 6 · Sacra; short quotation of the free summary
  • source Sierra blog · default tier 5 · Sierra; short quotation
  • source TechCrunch · default tier 7 · TechCrunch; short quotation
  • source Tech Startups · default tier 5 · TechStartups; short quotation
  • source 24/7 Wall St. · default tier 5 · 24/7 Wall St.; short quotation
  • source Yahoo Finance (syndicated Investing.com) · default tier 5 · Syndicated; short quotation
  • proxy capital_flow, market_structure · cadence irregular

The rule compares three readings from one compiler (Dealroom) for the largest application company, Anysphere (Cursor); moves under $300M are reporting noise. Other publishers reporting the same company (TechCrunch, Forbes) are the second source. Rising run-rate means the application layer is keeping more of the surplus.

Applied to dealroom.anysphere.revenue_run_rate_usd.pt.

Latest readings
  • 2026-06-05 anysphere · $4.0B⚑disputed: An annualised month reported by the press or the company, not booked or audited revenue, and often sourced to the company itself.run-rateas of 2026-06-05
  • 2026-04-30 anysphere · $3.0B⚑disputed: An annualised month reported by the press or the company, not booked or audited revenue, and often sourced to the company itself.run-rateas of 2026-04-30
Why this status, and its history

First reading. Dealroom reports Anysphere (Cursor) at an annualised run-rate of $3.0B at the end of April 2026 and $4.0B in early June 2026. Its direction rule compares three readings from the one compiler, so with two it reads emerging; the other application companies stay as evidence.

  1. 2026-09-27unmeasured to emergingconf — → 35 · Claude (model)

    First reading. Dealroom reports Anysphere (Cursor) at an annualised run-rate of $3.0B at the end of April 2026 and $4.0B in early June 2026. Its direction rule compares three readings from the one compiler, so with two it reads emerging; the other application companies stay as evidence.

    Weighed against it: One company, and the most successful one, so the series over-represents success; run-rates annualise one month and are not audited.

    sources12
Why this confidence

Rests on 41 observations from 24/7 Wall St., CNBC, Clay blog (GTM with Clay), Dealroom news, ElevenLabs blog, Forbes, Glean press releases, LawSites (LawNext), Legora newsroom, Lovable blog, Newcomer (free preview), PYMNTS, RuntimeWire, Sacra company pages (free summaries), Sierra blog, Tech Startups, TechCrunch, Yahoo Finance (syndicated Investing.com); best evidence tier 5, grade C; current.

Confidence is independent of status: 90–95 multiple strong independent sources; 70–89 good evidence, some ambiguity; 50–69 mixed or hard to operationalise; 0–49 limited or vague evidence.