Instructions to use bearzi/MiniMax-M2.7-JANG_2M with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use bearzi/MiniMax-M2.7-JANG_2M 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("bearzi/MiniMax-M2.7-JANG_2M")

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 bearzi/MiniMax-M2.7-JANG_2M with Pi:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "bearzi/MiniMax-M2.7-JANG_2M"

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": "bearzi/MiniMax-M2.7-JANG_2M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use bearzi/MiniMax-M2.7-JANG_2M 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 "bearzi/MiniMax-M2.7-JANG_2M"

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 bearzi/MiniMax-M2.7-JANG_2M

Run Hermes

hermes

MLX LM

How to use bearzi/MiniMax-M2.7-JANG_2M with MLX LM:

Generate or start a chat session

# Install MLX LM
uv tool install mlx-lm
# Interactive chat REPL
mlx_lm.chat --model "bearzi/MiniMax-M2.7-JANG_2M"

Run an OpenAI-compatible server

# Install MLX LM
uv tool install mlx-lm
# Start the server
mlx_lm.server --model "bearzi/MiniMax-M2.7-JANG_2M"
# Calling the OpenAI-compatible server with curl
curl -X POST "http://localhost:8000/v1/chat/completions" \
   -H "Content-Type: application/json" \
   --data '{
     "model": "bearzi/MiniMax-M2.7-JANG_2M",
     "messages": [
       {"role": "user", "content": "Hello"}
     ]
   }'

MiniMax-M2.7-JANG_2M

File size: 821 Bytes

916fe73

{
  "quantization": {
    "method": "jang-importance",
    "profile": "JANG_2M",
    "target_bits": 2,
    "actual_bits": 2.09,
    "block_size": 128,
    "calibration_method": "activations",
    "quantization_method": "mse-all",
    "scoring_method": "awq+hessian",
    "bit_widths_used": [
      2,
      4,
      8
    ],
    "quantization_scheme": "asymmetric",
    "quantization_backend": "mx.quantize",
    "hadamard_rotation": false
  },
  "source_model": {
    "name": "MiniMax-M2.7",
    "dtype": "bfloat16",
    "parameters": "227.6B"
  },
  "architecture": {
    "type": "moe",
    "attention": "gqa",
    "has_vision": false,
    "has_ssm": false,
    "has_moe": true
  },
  "runtime": {
    "total_weight_bytes": 59822505984,
    "total_weight_gb": 55.71
  },
  "format": "jang",
  "format_version": "2.0"
}