Skip to content
Slow Variables

Ask the data

Answers come from the same store as the site; every number is checked against the record it cites.

Model

Frontier-lab revenue run-rates

concentrating55/95 confidence, mixed or hard to operationalisegrade Ccoincident

1What this measures

Annualised revenue run-rates for OpenAI and Anthropic as compiled by Epoch AI from press reports; each point is an annualised month, not booked revenue.

Why it matters. Whether the model layer keeps surplus depends on revenue growing faster than the capital it consumes. Run-rate is the only public series, and it is press-sourced.

Proxy types
capital_flow, market_structure
Unit
USD
Cadence
monthly

2How we track this

  • series epoch.openai.revenue_run_rate_usd.pt
  • series epoch.anthropic.revenue_run_rate_usd.pt
  • series pymnts.openai.revenue_run_rate_usd.pt
  • series yahoo_finance.openai.revenue_run_rate_usd.pt
  • series cnbc.anthropic.revenue_run_rate_usd.pt
  • source CNBC · default tier 5 · CNBC; short quotation
  • source Epoch AI data hub — AI companies · default tier 5 · CC BY 4.0
  • source PYMNTS · default tier 5 · PYMNTS; short quotation
  • source Yahoo Finance (syndicated Investing.com) · default tier 5 · Syndicated; short quotation

Direction over 4 periods, dead-band $2.0B; higher = concentrating.

Four reports; moves under $2B are reporting noise. Rising run-rate concentrates revenue at the two frontier labs.

Applied to epoch.openai.revenue_run_rate_usd.pt.

3Tracker interpretation

Run-rates are compounding at rates no software company has printed, but every point is a leak to a journalist and definitions drift between gross and net.

4Evidence

Direction over the last 4 periods; the band is ±$2.0B around the window's first point. Higher reads concentrating.

Latest point
$40.0Bas of 2026-08-13
openai
Value the bands apply to
$40.0Bas of 2026-08-13

34 observations. Hollow points are disputed (see counterevidence). Every point links to its observation.

5Status and reasoning

concentratingsince 2026-09-10 · evaluate

Evaluator: 4e+10 (Epoch AI data hub — AI companies, 2026-08-13) reads concentrating over 4 periods with a dead band of 2e+09. Auto-reason; rule rationale: Four reports; moves under $2B are reporting noise. Rising run-rate concentrates revenue at the two frontier labs.

6Timeline notes

  • 2026-08-13 openai · $40.0Bas of 2026-08-13
  • 2026-02-28 openai · $25.0Bas of 2026-02-28
  • 2025-12-31 openai · $21.4Bas of 2025-12-31
  • 2025-12-18 openai · $19.0Bas of 2025-12-18
  • 2025-08-01 openai · $13.0Bas of 2025-08-01
  • 2025-07-30 openai · $12.0Bas of 2025-07-30

7Counterevidence

What cuts against this reading

Single-source press reports (tier 5); run-rate is annualised from one month; gross vs net of cloud revenue share is often unstated.

8Update history

  1. 2026-09-10unmeasured to concentratingconf — → 55 · evaluate

    Evaluator: 4e+10 (Epoch AI data hub — AI companies, 2026-08-13) reads concentrating over 4 periods with a dead band of 2e+09. Auto-reason; rule rationale: Four reports; moves under $2B are reporting noise. Rising run-rate concentrates revenue at the two frontier labs.

    obs:2932a2f9

9Confidence

55 / 95 — mixed or hard to operationalise

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

10Related