Harmonic, the AI startup co-founded by Robinhood CEO Vlad Tenev, has secured $120 million in fresh funding at a valuation of $1.45 billion. The investment, led by Ribbit Capital with support from Sequoia, Kleiner Perkins and new investor Emerson Collective, marks the company’s third major round in just 14 months. With this, the total funding raised by the company accounts for $295 million.
A new approach to reasoning
Unlike most generative systems trained to predict likely responses, Harmonic claims to solve the hallucination problem by transforming how AI reasons. Its core technology, called Mathematical Superintelligence (MSI), is built on formal logic rather than statistical guesswork. The company’s flagship model, Aristotle, is trained on synthetic math proofs and computer-generated problem-solving examples, enabling it to arrive at conclusions through verified logic rather than intuition.
Aristotle’s performance has already attracted attention beyond the investment world. In July, the model was tested at the International Mathematical Olympiad and achieved a top-tier score in line with leading research labs from Google and OpenAI.
According to CEO Tudor Achim, this accomplishment sparked a surge of investor interest.
AI built for precision
Harmonic’s technology operates by requiring its models to produce each step in their reasoning explicitly as computer code written in the Lean4 programming language. The code is then evaluated for precision, ensuring the final answer rests upon verifiable logic. This structure mirrors how mathematicians approach proofs rather than how conversational systems typically generate responses.
Processing this degree of rigour demands immense computing power, and much of the new funding will be devoted to training and scaling models capable of handling increasingly complex logic. Harmonic believes this foundation is critical for deploying AI in high-risk sectors such as aerospace and finance, where even small errors can have dangerous or costly consequences. For these industries, trust is paramount, and the company aims to make accuracy non-negotiable rather than aspirational.
Opening the gates for researchers
While Harmonic is not yet generating revenue, Aristotle is publicly available through a free API that developers and mathematicians can integrate into their own workflows. Researchers have been using it to evaluate challenging proofs and explore new theoretical breakthroughs that previously required months of manual effort.
Achim has indicated that commercialisation will come later, but for now, the priority is to cement the company’s position as the leader in logic-driven computing.
With investor confidence running high and demand for dependable AI intensifying, Harmonic has become one of the most closely watched contenders in the race to make AI more trustworthy by eliminating incorrect outputs often referred to as hallucinations.
