Bonsai 2 · 27B

A compressed 27B-class reasoning model with publisher-provided GGUF packs and a dedicated llama.cpp fork for CUDA, Metal, and CPU.

Verified sourceReleased Sep 17, 2026Source checked 9/23/2026

LOCALRENTED GPUOPEN WEIGHTS

At a glance

MY HARDWARE

Save your machine to see a personalized rating and its reasoning.

Private reasoning and coding on hardware that cannot hold conventional 27B FP16 weights.

Requires Prism ML’s modified llama.cpp runtime; stock builds do not support these packed weights.
The optional vision tower increases memory use for image input.

Use the publisher fork or its pinned binary. Stock llama.cpp cannot run the ternary packs correctly.

hf download prism-ml/Ternary-Bonsai-2-27B-gguf Ternary-Bonsai-2-27B-PQ2_0.gguf --local-dir .

REASONINGCODINGCOMPRESSED WEIGHTS

Use Bonsai 2’s dedicated low-bit runtime, download its PQ2_0 package, and verify a local response without accidentally using stock llama.cpp.

A multimodal reasoning model with a compressed key-value cache, published as open weights.

An experimental open-weight multimodal model with sparse attention and 262K native context.

An open-weight language model with switchable thinking and non-thinking modes.