Instructions to use prism-ml/Bonsai-27B-mlx-1bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use prism-ml/Bonsai-27B-mlx-1bit with MLX:

# Make sure mlx-lm is installed
# pip install --upgrade mlx-lm

# Generate text with mlx-lm
from mlx_lm import load, generate

model, tokenizer = load("prism-ml/Bonsai-27B-mlx-1bit")

prompt = "Write a story about Einstein"
messages = [{"role": "user", "content": prompt}]
prompt = tokenizer.apply_chat_template(
    messages, add_generation_prompt=True
)

text = generate(model, tokenizer, prompt=prompt, verbose=True)

Notebooks
Google Colab
Kaggle
Local Apps Settings
LM Studio

How to use prism-ml/Bonsai-27B-mlx-1bit with Pi:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "prism-ml/Bonsai-27B-mlx-1bit"

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "mlx-lm": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "prism-ml/Bonsai-27B-mlx-1bit"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

MLX LM

How to use prism-ml/Bonsai-27B-mlx-1bit with MLX LM:

Generate or start a chat session

# Install MLX LM
uv tool install mlx-lm
# Interactive chat REPL
mlx_lm.chat --model "prism-ml/Bonsai-27B-mlx-1bit"

Run an OpenAI-compatible server

# Install MLX LM
uv tool install mlx-lm
# Start the server
mlx_lm.server --model "prism-ml/Bonsai-27B-mlx-1bit"
# Calling the OpenAI-compatible server with curl
curl -X POST "http://localhost:8000/v1/chat/completions" \
   -H "Content-Type: application/json" \
   --data '{
     "model": "prism-ml/Bonsai-27B-mlx-1bit",
     "messages": [
       {"role": "user", "content": "Hello"}
     ]
   }'

Hermes Agent

How to use prism-ml/Bonsai-27B-mlx-1bit with Hermes Agent:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "prism-ml/Bonsai-27B-mlx-1bit"

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default prism-ml/Bonsai-27B-mlx-1bit

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use prism-ml/Bonsai-27B-mlx-1bit with OpenClaw:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "prism-ml/Bonsai-27B-mlx-1bit"

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "prism-ml/Bonsai-27B-mlx-1bit" \
  --custom-provider-id mlx-lm \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

Update README.md

by yogeshsaini1110 - opened Jul 19

base: refs/heads/main

←

from: refs/pr/5

Discussion Files changed

-4

Files changed (1) hide show

README.md +2 -4

README.md CHANGED Viewed

@@ -29,8 +29,7 @@ base_model:
 # 1-bit Bonsai 27B
-Full 27B-class reasoning in binary transformer weights — the first 27B-class model to run on a phone
 > **~14.2x** smaller than FP16 | **~90%** of FP16 intelligence retained | **~11 tok/s** on iPhone 17 Pro Max
 ## Highlights
@@ -81,8 +80,7 @@ Effective bits per weight: **1.125** (1 sign bit + 16-bit scale amortized over 1
 | FP16 (baseline)     | 16.0             | 54 GB      | 1.0x        |
 | **Binary g128**     | **1.125**        | **3.9 GB** | **14.2x**  |
-The binary build deploys at its native 1-bit g128 layout, so ideal and deployed sizes match. The deployed figure describes the language model alone — the only component that must stay resident for text inference; a negligible tail of normalization and scale parameters remains in higher precision.
 Unlike conventional low-bit builds — whose advertised labels understate their true average bit-width (a widely-used "2-bit" build of Qwen3.6-27B is really 2.8 bits/weight at 9.4 GB) — the Bonsai representation carries a bit-width that matches its name.
 ### Shipped Components

 # 1-bit Bonsai 27B
+Full 2
 > **~14.2x** smaller than FP16 | **~90%** of FP16 intelligence retained | **~11 tok/s** on iPhone 17 Pro Max
 ## Highlights
 | FP16 (baseline)     | 16.0             | 54 GB      | 1.0x        |
 | **Binary g128**     | **1.125**        | **3.9 GB** | **14.2x**  |
+The binary build deploys at its native 1-bit g128 layout, so ideal and deployed s
 Unlike conventional low-bit builds — whose advertised labels understate their true average bit-width (a widely-used "2-bit" build of Qwen3.6-27B is really 2.8 bits/weight at 9.4 GB) — the Bonsai representation carries a bit-width that matches its name.
 ### Shipped Components