OpenAI has acquired Glass Imaging, a Los Altos, California startup that uses neural networks to improve the images produced by small cameras, in a deal worth more than $300 million, The Wall Street Journal reported on September 14, 2026.
Neither company publicly confirmed the purchase or released terms, and OpenAI did not respond to requests for comment from TechCrunch. According to Pulse 2.0, the price represents a sharp increase on the roughly $100 million valuation Glass Imaging carried in a funding round the previous year.
OpenAI has not said what it will do with Glass Imaging's technology or team. The deal comes as the ChatGPT maker moves into consumer devices.
What Glass Imaging built
Ziv Attar and Tom Bishop founded Glass Imaging in 2019 and served as chief executive and chief technology officer. Both came from Apple, where they worked on the computational photography behind Portrait Mode. Attar joined Apple in 2015 when it bought his Israeli camera startup LinX Imaging for about $20 million, SiliconANGLE reported, citing Globes.
The company's product, GlassAI, works as a neural image signal processor. Instead of editing a photo after it is taken, it processes raw data from the camera sensor using a network trained on the specific flaws of one lens and sensor combination. Glass measures each module's optical behaviour, sensor response and noise profile, then builds a model around them. That means tuning is specific to each phone the software ships in.
Glass says the approach can improve camera performance by as much as ten times and recover detail that conventional processing misses, especially as pixels shrink. It has pitched the software for drones, wearables and other compact cameras as well as phones.
The technology had reached shipping products. Chinese smartphone maker Honor uses it for zoom on its 600 series, and a raw video version has been demonstrated on Qualcomm chipsets, according to SiliconANGLE.
How GlassAI works
Ordinary camera pipelines treat steps such as demosaicing, noise reduction, sharpening and frame fusion separately. GlassAI handles them in a single pass and models the optical behaviour of the individual lens, SiliconANGLE reported. The company says the result is detail the sensor actually captured rather than detail a model invented.
The approach is aimed in particular at very small pixels. Shivansh Rao of Glass Imaging's machine learning team told PetaPixel that submicron pixels encode fine detail in ways conventional processing cannot decode, and that in controlled tests the neural method improved a standard resolution measure by more than 50% as pixel size fell from 0.75 to 0.35 microns.
The founders and the funding
Glass Imaging raised about $29 million before the sale. Insight Partners led a $20 million Series A in May 2025, following a $9.3 million extended seed round in 2024, SiliconANGLE reported. Other backers named in coverage of the deal include GV, Alphabet's venture arm, Future Ventures and Abstract Ventures.
At more than $300 million, the reported price is roughly ten times the capital the company raised.
Why OpenAI is buying
OpenAI is widely reported to be working on its own hardware, including AI companion devices, earbuds and a screenless speaker. In 2025 it bought io Products, the device company co-founded by former Apple design chief Jony Ive, for $6.5 billion in stock, and it has since hired dozens of Apple employees.
Cameras are an important sensor for AI devices that need to understand their surroundings, read text or recognise objects. Glass Imaging gives OpenAI engineers who have taken computational photography from research into phones on sale, and software designed to get good images from small, constrained camera modules.
A busy year for OpenAI deals
Glass Imaging is the latest in a string of purchases by OpenAI in 2026, which also included the tech podcast company TBPN in April.
The company has not said whether Glass Imaging's existing licensing arrangements with phone makers such as Honor will continue.
