Bonsai 2 27B: Near-Lossless Compression in a 9x Smaller Footprint

PrismML releases Ternary Bonsai 2 27B, a compressed multimodal model based on Qwen3.8 27B. It uses ternary weights to achieve a 5.9GB footprint, retaining 98.2% of the original model's aggregate benchmark performance. The model supports a 262K context window and is available under an Apache 2.0 license for local deployment.