Skip to main content
All weights live under the prism-ml org on Hugging Face. Names follow one convention:
FamilyFormatRepository pattern
Bonsai (1-bit)GGUFprism-ml/Bonsai-{SIZE}-gguf
Bonsai (1-bit)MLXprism-ml/Bonsai-{SIZE}-mlx-1bit
Ternary (1.58-bit)GGUFprism-ml/Ternary-Bonsai-{SIZE}-gguf
Ternary (1.58-bit)MLXprism-ml/Ternary-Bonsai-{SIZE}-mlx-2bit
{SIZE} is 27B, 8B, 4B, or 1.7B. Not sure which variant you want? See the lineup for size trade-offs, and Formats & runtime support for GGUF vs MLX.

​
Automatic (demo repo)

setup.sh resolves the repository from BONSAI_MODEL and BONSAI_FAMILY and downloads into models/. To fetch additional variants later without re-running full setup: 
BONSAI_FAMILY=ternary BONSAI_MODEL=8B ./scripts/download_models.sh
BONSAI_FAMILY=all BONSAI_MODEL=all ./scripts/download_models.sh   # everything

​
Manual (Hugging Face CLI)

pip install -U "huggingface_hub[cli]"
hf download prism-ml/Bonsai-8B-gguf --local-dir ./Bonsai-8B-gguf
Swap in any repository from the tables below. No authentication is required; the repos are public.

​
Repositories and sizes

​
27B

All 27B repos also carry an mmproj file for vision (+0.9 GiB), needed for image input regardless of format.
FamilyFormatRepositoryOn disk (weights)
Bonsai (1-bit)GGUFprism-ml/Bonsai-27B-gguf3.53 GiB
Bonsai (1-bit)MLXprism-ml/Bonsai-27B-mlx-1bit3.92 GiB
TernaryGGUFprism-ml/Ternary-Bonsai-27B-gguf6.66 GiB
TernaryMLXprism-ml/Ternary-Bonsai-27B-mlx-2bit7.05 GiB

​
8B

FamilyFormatRepositoryOn disk
Bonsai (1-bit)GGUFprism-ml/Bonsai-8B-gguf1.16 GB
Bonsai (1-bit)MLXprism-ml/Bonsai-8B-mlx-1bit1.28 GB
TernaryGGUFprism-ml/Ternary-Bonsai-8B-gguf2.18 GB
TernaryMLXprism-ml/Ternary-Bonsai-8B-mlx-2bit2.30 GB

​
4B

FamilyFormatRepositoryOn disk
Bonsai (1-bit)GGUFprism-ml/Bonsai-4B-gguf0.57 GB
Bonsai (1-bit)MLXprism-ml/Bonsai-4B-mlx-1bit0.63 GB
TernaryGGUFprism-ml/Ternary-Bonsai-4B-gguf1.07 GB
TernaryMLXprism-ml/Ternary-Bonsai-4B-mlx-2bit1.13 GB

​
1.7B

FamilyFormatRepositoryOn disk
Bonsai (1-bit)GGUFprism-ml/Bonsai-1.7B-gguf0.25 GB
Bonsai (1-bit)MLXprism-ml/Bonsai-1.7B-mlx-1bit0.27 GB
TernaryGGUFprism-ml/Ternary-Bonsai-1.7B-gguf0.46 GB
TernaryMLXprism-ml/Ternary-Bonsai-1.7B-mlx-2bit0.48 GB

​
What’s inside a GGUF repo

You usually want exactly one file from each repo:
  • 1-bit repos: Bonsai-{SIZE}-Q1_0.gguf is the file to use.
  • Ternary repos: Ternary-Bonsai-{SIZE}-Q2_0.gguf is the default, plus the FP16 reference weights (large; only for comparison work).
  • 27B repos only: also grab the mmproj file alongside the quant file — it’s required for image input and isn’t optional the way the FP16 reference is.
To pull just the file you need: 
hf download prism-ml/Ternary-Bonsai-8B-gguf Ternary-Bonsai-8B-Q2_0.gguf --local-dir .
MLX repos are standard mlx-lm model directories (config, tokenizer, safetensors); download the whole repo.

​
Next steps

Check runtime support

Make sure your runtime has native kernels for the file you downloaded.

Run it

Start the local server and connect your tools.