# PrismML Compresses a 27-Billion-Parameter Model to Ternary Weights and Keeps 98.2 Percent of Its Scores

Bonsai 2 27B stores nearly every weight as minus one, zero or plus one, shrinking a roughly 54-gigabyte model to under nine gigabytes under an Apache 2.0 licence.

- Published: 2026-09-21T06:21:17.793Z
- Canonical: https://polylog.news/ai/2026-09-21/prismml-compresses-a-27-billion-parameter-model-to-ternary-w
- Publisher: Polylog (AI desk)
- Section: tech
- Sources: [Polylog editors](https://polylog.news), [SiliconANGLE](https://siliconangle.com/2026/09/18/prismml-launches-bonsai-2-27b-a-high-intelligence-ai-model-so-small-it-fits-on-consumer-hardware/), [arXiv](https://arxiv.org/abs/2609.20824)

PrismML, an American startup founded by a team out of Caltech, has released Bonsai 2 27B, a ternary-quantized version of Qwen3.8-27B. Nearly every weight is stored as one of three values, minus one, zero or plus one, which is about 1.585 bi…

This story is for subscribers. Read it in full at https://polylog.news/ai/2026-09-21/prismml-compresses-a-27-billion-parameter-model-to-ternary-w (subscription information: https://polylog.news/pricing).