Instructions to use spicyneuron/GLM-5.1-MLX-2.9bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use spicyneuron/GLM-5.1-MLX-2.9bit 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("spicyneuron/GLM-5.1-MLX-2.9bit")

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
LM Studio

Pi new

How to use spicyneuron/GLM-5.1-MLX-2.9bit with Pi:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "spicyneuron/GLM-5.1-MLX-2.9bit"

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "mlx-lm": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "spicyneuron/GLM-5.1-MLX-2.9bit"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use spicyneuron/GLM-5.1-MLX-2.9bit 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 "spicyneuron/GLM-5.1-MLX-2.9bit"

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 spicyneuron/GLM-5.1-MLX-2.9bit

Run Hermes

hermes

MLX LM

How to use spicyneuron/GLM-5.1-MLX-2.9bit with MLX LM:

Generate or start a chat session

# Install MLX LM
uv tool install mlx-lm
# Interactive chat REPL
mlx_lm.chat --model "spicyneuron/GLM-5.1-MLX-2.9bit"

Run an OpenAI-compatible server

# Install MLX LM
uv tool install mlx-lm
# Start the server
mlx_lm.server --model "spicyneuron/GLM-5.1-MLX-2.9bit"
# Calling the OpenAI-compatible server with curl
curl -X POST "http://localhost:8000/v1/chat/completions" \
   -H "Content-Type: application/json" \
   --data '{
     "model": "spicyneuron/GLM-5.1-MLX-2.9bit",
     "messages": [
       {"role": "user", "content": "Hello"}
     ]
   }'

spicyneuron commited on Apr 26

Commit

cefeab8

verified ·

1 Parent(s): cea20af

Update README.md

Browse files

Files changed (1) hide show

README.md +17 -9

README.md CHANGED Viewed

@@ -10,7 +10,13 @@ tags:
 - mlx
 ---
-[GLM 5.1](https://huggingface.co/zai-org/GLM-5.1) optimized for MLX. A mixed-precision quant that balances speed, memory, and accuracy.
 # Usage
@@ -22,14 +28,6 @@ uvx --from mlx-lm mlx_lm.server \
   --model spicyneuron/GLM-5.1-MLX-2.9bit
 ```
-# Methodology
-Quantized with a [mlx-lm fork](https://github.com/ml-explore/mlx-lm/pull/922), drawing inspiration from Unsloth/AesSedai/ubergarm style mixed-precision GGUFs.
-MLX quantization options differ than llama.cpp, but the principles are the same:
-- Sensitive layers like MoE routing, attention, and output embeddings get higher precision
-- More tolerant layers like MoE experts get lower precision
 # Benchmarks
 metric | baa-ai/GLM-5.1-RAM-270GB-MLX | 2.9bit (this model)
@@ -52,3 +50,13 @@ mlx_lm.evaluate --tasks hellaswag --seed 123 --num-shots 0 --limit 2000
 mlx_lm.evaluate --tasks piqa --seed 123 --num-shots 0 --limit 2000
 mlx_lm.evaluate --tasks winogrande --seed 123 --num-shots 0 --limit 2000
 ```

 - mlx
 ---
+[GLM 5.1](https://huggingface.co/zai-org/GLM-5.1) optimized to run _comfortably_
+on a Mac Studio M3 512. This is the smaller, compact version. Quality-first
+version [here](https://huggingface.co/spicyneuron/GLM-5.1-MLX-3.6bit).
+- A mixed-precision quant that balances speed, memory, and accuracy.
+- 3-bit baseline with important layers at 4, 8 and BF16.
+- Fits into ~280 GB memory, leaving plenty of room to run parallel models (ex: Minimax M2.7, Qwen 3.6 35B).
 # Usage
   --model spicyneuron/GLM-5.1-MLX-2.9bit
 ```
 # Benchmarks
 metric | baa-ai/GLM-5.1-RAM-270GB-MLX | 2.9bit (this model)
 mlx_lm.evaluate --tasks piqa --seed 123 --num-shots 0 --limit 2000
 mlx_lm.evaluate --tasks winogrande --seed 123 --num-shots 0 --limit 2000
 ```
+# Methodology
+Quantized with a [mlx-lm fork](https://github.com/ml-explore/mlx-lm/pull/922),
+drawing inspiration from Unsloth/AesSedai/ubergarm style mixed-precision GGUFs.
+MLX quantization options differ from llama.cpp, but the principles are the
+same:
+- Sensitive layers like MoE routing, attention, and output embeddings get higher precision
+- More tolerant layers like MoE experts get lower precision