Axiom announces complete Lean auto-formalization of the Classification of Finite Simple Groups in 24 days, producing roughly 40 million lines of Lean code.
Axiom AI Formalizes Longest Math Proof in Lean in 24 Days
Formal verification removes human referee fallibility from mathematical proof-checking by having a trusted kernel program verify each line. A proof of this scale being machine-checked addresses a long-standing gap between AI performance on short competition problems and real extended mathematics.
When stories like this move, I write about them in an email each morning.
One free email each morning, around 8 AM US Eastern.
Check your email.
I have sent you a confirmation link. One click and you are in.
Not in your inbox? Look in spam or promotions. If it's there, mark it "Not spam" so future editions land in your inbox.
The full picture
Axiom announced a complete Lean auto-formalization of the Classification of Finite Simple Groups, described as the longest proof in mathematics at roughly 15,000 pages spanning more than 500 papers and over 100 authors across generations. Axiom's swarm of AI agents, guided by mathematicians, produced approximately 40 million lines of Lean code in 24 days. The Classification had no single person who had read all of it. Earlier in 2026, Axiom had formalized the 246 Prime-Gaps Theorem in Lean 4 using its multi-agent system AxiomProver, which operates through a three-stage pipeline: a blueprint phase labeling every definition, lemma, and theorem with dependencies; proof generation by AxiomProver; and verification against Mathlib. Axiom attributes the capability jump from the Prime-Gaps work to the Classification formalization to simultaneous improvements across models, the agent harness, and infrastructure. Earlier this year, getting AxiomProver to write a correct 10,000-line Lean proof took weeks; the Classification task was about 4,000 times larger.
How it developed
Axiom Math formalizes the 246 Prime-Gaps Theorem in Lean 4 using AxiomProver.
Sources
- Anthropic's Fermat's Last Theorem formalization claim
- AI proofs of Navier-Stokes and Yang-Mills
- Google Cogentic and Meta Muse Spark proofs
- OpenAI's internal Astra model produced Lean 4-verified proofs for ten long-standing open mathematics problems
BackgroundAI agents: models that take actions
Get the daily email
When stories like this move, I write about them in an email each morning.
One free email each morning, around 8 AM US Eastern.
Check your email.
I have sent you a confirmation link. One click and you are in.
Not in your inbox? Look in spam or promotions. If it's there, mark it "Not spam" so future editions land in your inbox.