How AI Companies Are Valued at 100x Revenue in 2026

Key Takeaways

• Harvey, Legora, and Sierra each crossed $100m ARR within nine months, yet received valuations ranging from 50x to 100x revenue
• Valuation multiples vary dramatically (25-125x) despite similar growth trajectories
• Revenue growth rate alone does not explain premium valuations; market position matters more
• Multiples accelerated in early 2026, returning to 100x ARR levels last seen in 2021—but with 3x faster underlying growth

The AI Valuation Puzzle

Three major AI companies announced crossing $100m in annual recurring revenue within nine months of each other: Harvey (August 2025), Sierra (November 2025), and Legora (April 2026). Yet their valuations tell a surprising story.

Harvey reached $5b post-money valuation at roughly 67x revenue. Sierra commanded $10b at 100x+ revenue. Legora captured $5.6b at 56x revenue. Meanwhile, Ramp and Decagon bookended the group at $1.4b and $35m, representing the broader valuation spectrum for AI businesses.

The puzzle: Growth rate doesn't explain the gap. The fastest-growing company was priced near the bottom of the range. Instead, market category and competitive positioning appear to drive the premium.

Why Multiples Don't Compress at Scale

Typically, valuation multiples compress as companies scale—higher revenue should command lower multiples. Yet these AI companies trade within just one or two consistent valuation bands despite significant growth in ARR.

In January 2026, multiples accelerated further for Legora, Sierra, and Ramp. This reflects both sustained acceleration in revenue growth and a more favorable fundraising market for AI-category leaders. The pattern mirrors historical examples like Databricks, where premium multiples persisted through scaling.

The Return of 100x ARR Valuation

In 2021, the market valued select private startups and public companies at 100x annual recurring revenue—a premium tier reserved for category leaders. Five years later, we're seeing those valuation levels again, but with a critical difference: the underlying growth is approximately 3x faster.

This suggests the market isn't paying for current revenue alone. It's pricing in the pace of expansion, market capture potential, and the rarity of AI businesses operating at scale.

Conclusion

The AI valuation premium isn't random. Harvey, Legora, and Sierra demonstrate that market position and growth trajectory determine pricing more than absolute revenue size. As these companies accelerate through 2026, watch whether multiples stabilize, compress, or climb further—signaling whether the market's confidence in AI-driven growth remains justified.

---

Original source: AI Harness' ARR Multiples

powered by osmu.app