This could be a big deal. Sure, you can run Qwen3.8 27B locally but if you do, practically nothing else can happen on your computer, so we need smaller models like this one to make local inference practical
PrismML
@PrismML
Sep 17
Today, we’re announcing Ternary Bonsai 2 27B. Based on Qwen3.8 27B, Bonsai 2 27B is 9x smaller than its full-precision counterpart while retaining 98.2% of its aggregate benchmark performance. Two months after the first Bonsai 27B release, the biggest change is quality. The footprint remains 5.9 GB, but the gap to full precision has narrowed materially, with particularly strong gains in agentic coding, multimodal reasoning, and long-horizon tool use. Ternary Bonsai 2 27B is available today under Apache 2.0.

Sep 19, 2026 · 5:42 AM UTC

2
1
23
3,041
RelevantRecentLikes
local models get interesting when they leave enough machine for the interface
13
Smaller models make local inference practical only if the full loop is efficient: quantization, latency, memory use, and task-specific evals matter as much as benchmark retention. The payoff is a system that can be private, cheap, and responsive by default.
10