Mini tools1 Bit Bonsai 1.7B
By PrismML
1-bit Bonsai 1.7B is PrismML’s smallest public Bonsai model and the most aggressively optimized for mobile-scale deployment. PrismML says it needs only 0.24GB of memory and can reach 130 tokens per second on an iPhone 17 Pro Max, highlighting very high energy efficiency and solid accuracy despite its size. It is positioned as a genuinely on-device model for smartphones and other constrained hardware, where minimal memory use, low power draw, and fast response matter more than maximum reasoning depth or larger-model breadth.
Text
Overview
1-bit Bonsai 1.7B is PrismML’s smallest Bonsai model, built for extremely fast on-device inference. With a memory footprint of just 0.24GB, PrismML says it can reach 130 tokens per second on an iPhone 17 Pro Max, making it a lightweight edge model aimed at mobile and highly power-constrained applications.
About PrismML
Industry:
Artificial Intelligence
Company Size:
6
Location:
Pasadena, California, US
Website:
prismml.com
Last updated: July 7, 2026
