APIUp to 28% cheaper than official pricesTry the API →
Hermes
APIModelsCases
Back to News
modelLlmsGpus

Prism ML's Ternary Bonsai 2 Squeezes a 27B Reasoning Model Into 8.6 GB

Prism ML's Bonsai 2 27B compresses a 27B reasoning model to 8.6 GB using ternary weights, retaining 98.2% of FP16 benchmark quality.

Prism ML's Ternary Bonsai 2 Squeezes a 27B Reasoning Model Into 8.6 GB
Source
PrismML
Published
Author
AlphaSignal Newsroom
Read
1 min read

Prism ML's Bonsai 2 27B compresses a 27B reasoning model to 8.6 GB using ternary weights, retaining 98.2% of FP16 benchmark quality.

Reporting is indexed from AlphaSignal. Rights remain with the original publisher and cited sources.

Read original report

Next reads

Swift 1.5 Heretic Build Cuts Refusals but Slows Decoding Speed by 28%
AlphaSignal · model

Swift 1.5 Heretic Build Cuts Refusals but Slows Decoding Speed by 28%

Google Runs Gemma 4 on a TPU in Orbit for the First Time
Google Gemma · news

Google Runs Gemma 4 on a TPU in Orbit for the First Time

NVIDIA Shrinks DeepSeek V4.1 Flash by 75 GiB With Almost No Quality Loss
NVIDIA AI · model

NVIDIA Shrinks DeepSeek V4.1 Flash by 75 GiB With Almost No Quality Loss