Harmonic, an AI startup co-founded by Robinhood CEO Vlad Tenev, announced on July 28 the beta launch of its AI chatbot app for iOS and Android platforms. The core AI model behind the app is called Aristotle, and it promises users accurate and error-free answers related to mathematical reasoning and calculations.
Harmonic aims to create “mathematical superintelligence” and intends to support users in fields such as mathematics, physics, statistics, and computer science through the Aristotle model.
The company’s CEO, Tudor Achim, stated that Aristotle formally verifies the correctness of its answers to mathematical problems, providing “hallucination-free” results. This feature distinguishes it from current AI models, which often produce errors and inaccurate information (hallucinations).
Aristotle achieved gold medal-level performance on official tests at the 2025 International Mathematical Olympiad (IMO). Although AI models developed by Google and OpenAI also showed strong results this year, they completed the tests informally in natural language rather than through formal machine-readable formats.
Harmonic plans to soon release an API for enterprises and a web app for general users to access Aristotle.
In June, Harmonic raised $100 million in funding, reaching a valuation of $875 million. CEO Tudor Achim noted that the company is rapidly advancing toward its goal of mathematical superintelligence and that investors consider the valuation appropriate given the startup’s broad ambitions.
Experts highly value AI’s achievements in mathematics, seeing them as both a technological breakthrough and a gateway to applications in other domains. Success in mathematics—a precise and verifiable field—demonstrates core logical capabilities of AI.
Achim added that Aristotle produces its answers in the open-source programming language Lean, and before delivering responses, they undergo an algorithmic verification process. This helps prevent AI hallucinations. Similar verification technologies are used in high-risk industries like medical devices and aviation.
Although achieving hallucination-free performance in AI models remains very challenging, Harmonic’s Aristotle represents significant progress in this area. Meanwhile, hallucinations remain a major issue for many leading AI models.
