| Company ▼ | Category ▼ | Layer ▼ | Founded ▼ | Valuation ▼ | ARR ($M) ▼ |
|---|
Every ARR number tells a story. Click any company to decode the economics — from GPU capex to labor displacement.
| Company ▼ | Category ▼ | Layer ▼ | Founded ▼ | Valuation ▼ | ARR ($M) ▼ |
|---|
ARR figures compiled from public earnings reports, fundraising announcements (press releases, SEC filings), and credible third-party estimates (The Information, PitchBook, CB Insights). Valuations reflect most recent known funding rounds. Last updated March 2026.
Token throughput derived from ARR ÷ blended token price. GPU counts assume H100-equivalent at 150 tok/s baseline. Inference cost = GPUs × $90K/yr. Power draw assumes 700W per GPU. All parameters are adjustable via interactive sliders in the detail panel.
Seat counts derived from ARR ÷ (monthly price × 12). FTE equivalents based on hours replaced per seat per day ÷ 2,080 annual work hours. Human salary benchmarks sourced from Bureau of Labor Statistics and Glassdoor. ROI = total labor value displaced ÷ total AI spend.
All calculations are illustrative models, not financial advice. Actual GPU utilization rates, pricing tiers, enterprise discounts, and seat counts vary significantly. Companies with non-standard pricing (usage-based, per-project) are flagged separately. This analysis covers a curated subset of the AI economy, not an exhaustive census.