Keyboard shortcuts

Press or to navigate between chapters

Press S or / to search in the book

Press ? to show this help

Press Esc to hide this help

Lean 4 beyond research

A snapshot, taken at the start of the Fall 2026 semester, of organizations using Lean 4 outside of research: what each one is doing with it, and the scale of the organization doing it. Every figure and claim is cited on the snapshot page, whose numbers link through to all thirty sources.

OrganizationValuationLean 4 workEmphasis
Google DeepMind / AlphabetAlphabet $4.13TAlphaProof — reinforcement learning over Lean; olympiad-level formal reasoning, published in NatureAutoformalization
Microsoft$3.69TOriginated Lean. Aeneas-based Lean verification of Rust SymCrypt, with ML-KEM and SHA-3 proofs shipping in Windows Insider buildsVerified production
Amazon / AWS$2.75TCedar authorization language, modelled in Lean and differentially tested against the production Rust; the proofs found 4 validator bugs and testing found 21 moreVerified production
Anthropic$965BClaude raised a lower bound on zeta zeros from 41.6% to 67.2%, with a Lean formalization that passes the standard validation toolNew mathematics
OpenAI$852BTen decade-open problems, each shipped with a machine-checkable Lean 4 certificateNew mathematics
ByteDance>$600BBFS-Prover — best-first search over Lean 4, open-sourced; 72.95% on MiniF2FTheorem proving
Tencent$503BHunyuanProver — data synthesis at scale with guided tree searchTheorem proving
DeepSeek~$74BDeepSeek-Prover-V2 — subgoal decomposition by reinforcement learning; 88.9% on MiniF2F-testTheorem proving
Moonshot AI~$50BKimina-Prover, built with Project NuminaTheorem proving
Mistral AI~$23BLeanstral — the first open-source code agent designed for Lean 4, Apache 2.0Autoformalization
Harmonic$1.45BAristotle — formally verified gold-medal-level performance at IMO 2025Autoformalization
Math, Inc.UndisclosedGauss completed the Tao–Kontorovich challenge to formalize the strong Prime Number Theorem in Lean; OpenGauss released MIT-licensedAutoformalization
HuaweiEmployee-ownedMathesis — natural language to Lean 4 via an RL-trained autoformalizerAutoformalization
NethermindPrivateEVM and Yul semantics in Lean, passing 99.99% (22,330/22,332) of the Cancun execution tests; Halva found a Keccak-256 bug in Scroll’s circuitVerified production
GaloisPrivateFVSpec — 2,772 property-based tests translated into 9,415 Lean 4 specificationsVerified production

Market capitalizations retrieved 2 September 2026, in USD; private valuations are the most recent round, and some were still open at that date. Each claim above is cited on the snapshot page, whose figures link through to all thirty sources.