The brief

Analyst view

• Used by 80% of the Am Law 100 and five of the Fortune 10, per Harvey.
• Tenet (Aug 2026), its first own model, post-trains Kimi K3 for long legal tasks and nearly doubles completion on LAB.
• Valuation rose from $8B (Dec 2025) to $11B (Mar 2026) to $15.5B (Sept 2026).

Technical approach

As reported

Routing

Best model per task

Harvey routes work to models from OpenAI, Anthropic, Google and now its own, picking whichever suits each legal task.

Agents

Multi-step legal work

Agents break a matter into steps, read documents, use tools and produce a memo, marked-up contract or diligence summary.

Evaluation

Benchmarks written by lawyers

LAB grades agents on 1,200+ realistic tasks against 75,000+ lawyer-written checks; a task passes only if all checks do.