OpenAI has acquired Torch, a small health-focused startup, and its entire team will now join the company to work on health and wellness features for ChatGPT. The update was shared by Torch co-founder Ilya Abyzov on January 13.
Torch was founded in 2024 by Abyzov, who earlier co-founded Forward Health. The startup was built by a compact four-member team that included Eugene Huang, James Hamlin and Ryan Oman. With the acquisition, all four will move to OpenAI.
While the companies did not officially reveal financial details, a person familiar with the deal said OpenAI paid around $100 million in equity for Torch.
Abyzov said the team’s focus will be on developing ChatGPT Health into a practical and reliable AI-based health tool. He pointed to ChatGPT’s massive user base as a key reason behind the move, noting that millions already use the platform for health-related questions every week.
Torch was created to solve a common problem in healthcare: scattered medical data. The platform was designed to bring information from hospitals, diagnostic labs, wearable devices and consumer health services into one place, making it easier for AI systems to understand a person’s overall health picture.
According to Abyzov, early users found value in having their health data organised in a single system, which helped them make better sense of reports, trends and medical history.
He also addressed concerns around privacy and safety, saying the team would not have joined OpenAI without confidence in its approach to protecting sensitive data. Collaboration with doctors and a high standard of product quality, he said, were also non-negotiable.
Before Torch, the same team worked together at Forward, where they aimed to build healthcare services at scale. While Torch took a different path than originally planned, Abyzov said becoming part of OpenAI brings that long-term vision closer.
In a separate blog post, Torch explained that AI struggles to deliver meaningful health insights when medical data is spread across multiple hospitals, apps and portals. Bringing everything together, the team believes, is essential for AI to truly help people manage their health better.
