Bonsai 2.

The small-tree idea made literal: ternary weights squeeze a vision-and-reasoning family into downloads that leave more of your device free for the rest of your day.

View on Hugging Face See LocalEngine variants ↓

Model makerPrismML

Supported in LocalEnginemacOS

Upstream licenseApache-2.0

SourceBonsai 2 collection ↗

Where it earns its space

Built for a real local rhythm.

These are good-fit starting points—not promises of benchmark performance. Choose the smallest listed option that comfortably covers your work.

Image-aware chat with a smaller desktop download

Longer private reasoning on capable Macs

A compact Mac vision assistant

In LocalEngine

Configured variants, by device.

The formats below are the downloads currently represented in LocalEngine’s model catalogs. GGUF runs through the llama.cpp-based runtime; MLX options use the Apple Silicon MLX runtime.

macOS

27B

Formats: Ternary Q2_0 GGUF + Q8_0 vision projector
Local note: The 27B build is about 7.6 GB and needs a Mac with 16 GB of memory or more.

About the source

Model identity, task category, supported formats, and license are based on the linked Hugging Face Bonsai 2 collection and the LocalEngine macOS/iOS catalogs. Hardware guidance reflects the packaged variants, not a universal system requirement.