See the latest news and media coverage for PrismML. We track all announcements, press releases, and industry mentions in real time, all in one place.
Quantized AI model developer
prismml.comLast updated
In short: PrismML emerged from stealth with 1-bit models, launched Bonsai 27B, and drew Apple's interest in on-device iPhone AI.
The update optimizes the 2-billion-parameter model for Qualcomm's AR1 Gen 1 platform, enabling richer multimodal experiences with significantly reduced memory usage and increased speed.
It compresses a 27B-parameter multimodal model to 3.9GB, enabling local reasoning, coding, and agentic workflows on devices.
The compact models run high-quality diffusion inference on local devices like phones and laptops.
The models use a 1.58-bit weight architecture for high efficiency and are available in 8B, 4B, and 1.7B sizes.
According to The Information, Apple may be interested in the technology of startup company PrismML. The company focuses on shrinking AI models that originally required servers to operate, allowing them to run on devices while maintaining similar levels of intelligence. This technology is expected to enable iPhones to run larger-scale AI models locally, reducing reliance on cloud processing. Details of PrismML's Technology: The report states that PrismML has successfully taken Alibaba's open-source large language model Qwen 3.6...
A small start-up just released a 27-billion-parameter AI model that can be run for free on an iPhone.
Apple is testing PrismML technology that could run larger AI models directly on future iPhones, improving speed, privacy, and ...
MacKenzie Sigalos / CNBC: PrismML launches Bonsai 27B, a model based on Qwen3.6 27B that it says runs natively on Apple devices via MLX; its...
Track PrismML and your other target companies to get real-time alerts and weekly summaries delivered straight to your inbox.
Browse news for competitors to PrismML and other trending companies.