Keenable in September 2028: most likely $24M revenue and a $312M valuation, but the credible range spans $6M–$70M
Asked:
predict keenable.ai's revenue and valuation 2 years from now.
A scenario forecast for September 7, 2028, built from three weighted cases — 25% downside, 50% base, 25% upside — over annualized revenue and implied valuation in USD millions. Keenable disclosed no current revenue, ARR, or valuation, so the model rests on its $26M Accel-led seed, a 100B+ document web index, production API use at several AI labs, and public $1 per 1,000-request pricing; the probability-weighted estimate lands at $31M revenue and $483M valuation. These are forecasts, not company guidance.
Bars span the downside-to-upside range on log scales; dot size reflects scenario probability. Amber marks the probability-weighted mean, which sits above the base case because the upside is far larger than the downside. Hover any dot for the case behind it.
Valuation uncertainty is wider than revenue uncertainty: the 8x–18x multiple range turns a ~12x revenue spread into a 26x valuation spread, from $48M to $1.26B —
cloudedjudgement.substack.comThe probability-weighted mean ($31M / $483M) exceeds the most-likely base case ($24M / $312M) — the distribution is right-skewed, so plan on the base and treat the mean as an expected value —
cryptobriefing.comAt the public $1 per 1,000-request entry price, the base case implies ~65.8M requests a day and the upside ~191.8M — a sanity check, not a pricing claim —
keenable.aiThe base case assumes today's production users — several AI labs and inference providers reported by TechCrunch — expand, with enterprise contracts supplementing metered API revenue —
techcrunch.com
Scenario table
| Scenario | Prob. | Revenue $M | Valuation $M | Multiple | Req./day (M) | Operating case | Source |
|---|
Assumptions and method
Valuation equals forecast annualized revenue times an assumed multiple: 8x downside, 13x base, 18x upside. The cited 2026 high-growth software median is 10.8x; the base and upside apply a premium for strategic AI-search infrastructure, while the downside discounts for customer concentration and execution risk. Request-equivalent volume divides revenue by the public $1 per 1,000-request entry price and 365 days; realized pricing will differ with enterprise contracts, volume tiers, content-fetch charges, and product mix. Key uncertainties: customer concentration, inference and search gross margins, hyperscaler competition, web-content licensing, price compression, and whether independent search infrastructure becomes a standard layer for agents. The wide range is candid: with no disclosed revenue or valuation, only scenarios are defensible.
Scenario forecast dataset: 4 rows (3 scenarios + 1 probability-weighted summary) dated 2028-09-07, built September 2026 from four independent public sources. Revenue is annualized run-rate and valuation is implied enterprise value, both in USD millions; request equivalents are millions per day at the $1/1,000-request entry price. No disclosed financials existed to extrapolate; nothing was cut.