PrismML releases Bonsai 27B: 1-bit Qwen3.6-27B for laptops and phones

Bonsai 27B ships under Apache 2.0 in two builds: a ternary variant at 1.71 bits per weight (~5.9GB) and a 1-bit "binary g128" variant (~3.9GB), down from ~54GB. PrismML says the 1-bit build keeps ~90% of benchmark scores, fitting a 27B model on iPhones.
How this story unfolded
6 days · 1 report · 4 community posts · from Jul 14
Daily brief
Get tomorrow's AI brief in your inbox
More stories today
- Sequoia Capital invests in AI-native video platform Preview
- US Launches Effort to Speed Trade in AI Goods Between Allies
- DeepMind launches SL2T sign language-to-text model
- Liquid AI releases LFM2.5-VL-3B vision-language model for edge
- Grok and Meta's release discussed on ETN podcast episode