OpenAI buys four-employee health startup Torch in $100 million equity deal
OpenAI has acquired Torch, a one year old healthcare startup with just four employees, in a deal valued at around $100 million. The team will now join OpenAI to work on ChatGPT Health.
New Delhi: Right after announcing ChatGPT health, OpenAI has quietly made another move into healthcare, this time by buying a very small startup with a very specific idea. The company confirmed it has acquired Torch, a one year old health tech startup with just four employees, in a deal that values the company at around $100 million in equity, as first reported by The Information.
The acquisition did not come with a flashy announcement. Still, the direction is clear. OpenAI is building out ChatGPT Health, and Torch’s work fits neatly into that plan. The entire Torch team will now join OpenAI, giving the AI firm a ready made group focused only on health data and memory.
What Torch was building before the acquisition
Torch was working on a consumer health app that tried to pull together a person’s medical data from many places. That includes doctor visits, lab results, wearables, medical imaging, and wellness tests. Anyone who has chased PDFs and hospital portals will get the idea.
The Torch team described its system as "a medical memory for AI, unifying scattered records into a context engine.” The goal was simple. Give AI a long term view of a person’s health, instead of one off answers.
Torch co founder Ilya Abyzov said the team first met while working at Forward Health. Forward ran AI powered doctors’ offices and raised more than $400 million before shutting down in late 2024. Torch survived that collapse. Now it has landed inside OpenAI.
Why OpenAI wanted a four person startup
This was an acqui-hire. OpenAI is not buying scale. It is buying focus. Torch already relied on OpenAI models for parts of its product, which made integration easier.
The team and its technology will now become part of ChatGPT Health. OpenAI recently announced the service for users who want help analyzing and managing their health through ChatGPT.
From a user point of view, this could mean fewer repeated questions. I have typed the same health details into apps more times than I can count. A system that remembers context sounds useful, even if it raises new privacy questions.