Harmonic’s AI reasoning model Aristotle has demonstrated gold medal performance by solving five out of six problems at the 2025 International Mathematical Olympiad, producing formally verified proofs for each solution. The model uses the Lean theorem prover to generate step-by-step logical verifications that a computer can independently check, a major achievement in mathematical AI.
Verified Proofs Over Simple Answers
Unlike traditional approaches that only deliver correct answers, Aristotle’s use of Lean ensures that every step of its reasoning process meets strict logical validation, eliminating uncertainty about solution correctness. This method represents a significant advance beyond previous AI efforts at the IMO, which Harmonic describes as relying on less rigorous solution verification standards.
Aristotle’s capabilities extend beyond competition problems: it also solved a variant of Erdős Problem #124, producing a verified proof within about one minute. This demonstrates the model’s potential for tackling a broad range of complex mathematical challenges.
Harmonic, co-founded by Robinhood CEO Vlad Tenev and Tudor Achim, released a beta chatbot app giving iOS and Android users direct access to Aristotle’s reasoning powers. Despite Tenev’s involvement, the company clarifies that Aristotle and Harmonic operate independently from Robinhood’s business, including its cryptocurrency ventures.
For investors in crypto and AI, the connection to Tenev has sparked speculation, but there is no blockchain aspect to Harmonic’s technology nor plans to integrate it with Robinhood’s digital assets. Thus any tokens linked to the news should be approached cautiously.
