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

Libraries

How to use JANGQ-AI/MiniMax-M2.7-JANGTQ 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("JANGQ-AI/MiniMax-M2.7-JANGTQ")

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 JANGQ-AI/MiniMax-M2.7-JANGTQ with Pi:

Start the MLX server

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

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": "JANGQ-AI/MiniMax-M2.7-JANGTQ"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use JANGQ-AI/MiniMax-M2.7-JANGTQ 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 "JANGQ-AI/MiniMax-M2.7-JANGTQ"

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 JANGQ-AI/MiniMax-M2.7-JANGTQ

Run Hermes

hermes

OpenClaw new

How to use JANGQ-AI/MiniMax-M2.7-JANGTQ with OpenClaw:

Start the MLX server

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

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 "JANGQ-AI/MiniMax-M2.7-JANGTQ" \
  --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"

MLX LM

How to use JANGQ-AI/MiniMax-M2.7-JANGTQ with MLX LM:

Generate or start a chat session

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

Run an OpenAI-compatible server

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

dealignai commited on Apr 14

Commit

3bd8df6

verified ·

1 Parent(s): a9a574c

Rename JANGTQ_2L → JANGTQ in README

Browse files

Files changed (1) hide show

README.md +9 -9

README.md CHANGED Viewed

@@ -25,7 +25,7 @@ base_model_relation: quantized
 <div align="center">
-# MiniMax-M2.7 JANGTQ_2L
 **MiniMax M2.7 228B MoE — 2.15-bit codebook + Hadamard, 56.5 GB**
@@ -51,7 +51,7 @@ Hadamard-rotated input (QuIP# "rotate-input-once" math).
 Result: **smaller than affine 2-bit, higher quality than affine 2-bit, runs at
 89% of affine 2-bit speed** on Apple Silicon.
-| | JANG_2L (affine) | **JANGTQ_2L** | Δ |
 |---|---|---|---|
 | Disk size | ~63 GB | **56.5 GB** | **−10%** |
 | GPU memory | ~62.6 GB | **56.5 GB** | **−10%** |
@@ -74,7 +74,7 @@ faithfully.
 Tested 2026-04-13 on Mac Studio M3 Ultra. Reasoning enabled (MiniMax M2.7 is
 an always-reasoning model); `<think>…</think>` stripped before scoring.
-| Subject | JANGTQ_2L | JANG_2L (affine) | JANG_3L/4M |
 |---|---|---|---|
 | **astronomy** | **20/20 (100%)** | — | — |
 | **high_school_biology** | **20/20 (100%)** | — | — |
@@ -88,7 +88,7 @@ an always-reasoning model); `<think>…</think>` stripped before scoring.
 | logical_fallacies | 16/20 (80%) | — | — |
 | **Total** | **183/200 = 91.5%** | **~88%** | **~95.5%** |
-JANGTQ_2L sits cleanly between affine JANG_2L (88%) and the larger JANG_3L/4M
 (95.5%) — capturing most of the quality of the 3L/4M profiles at ~55-60% of
 their disk footprint.
@@ -133,7 +133,7 @@ Strip `<think>…</think>` from the response before using the final answer.
 | Architecture | MoE (256 experts, top-8 active), standard Q/K/V attention, partial RoPE |
 | Total parameters | 228.7 B |
 | Active per token | ~1.4 B |
-| Profile | **JANGTQ_2L** |
 | Format | **JANGTQ (codebook+Hadamard)** — `weight_format: mxtq` in `jang_config.json` |
 | Avg bits/param | ~2.15 |
 | Disk | **56.55 GB** |
@@ -143,7 +143,7 @@ Strip `<think>…</think>` from the response before using the final answer.
 | Context | 192 K tokens |
 | Chat template | Always-reasoning (`<think>\n` opened at assistant start) |
-## JANGTQ_2L Bit Allocation
 | Component | Bits | Format | Why |
 |---|---|---|---|
@@ -178,7 +178,7 @@ from huggingface_hub import snapshot_download
 from jang_tools.load_jangtq import load_jangtq_model
 from mlx_lm import generate
-model_path = snapshot_download("JANGQ-AI/MiniMax-M2.7-JANGTQ_2L")
 model, tokenizer = load_jangtq_model(model_path)
 messages = [{"role": "user", "content": "Explain photosynthesis in 5 sentences."}]
@@ -193,7 +193,7 @@ print(out)
 On first load you'll see log lines like:
 ```
-Loading JANGTQ: MiniMax-M2.7-JANGTQ_2L
   seed=42, bits_map={'attention': 8, ..., 'routed_expert': 2, ...}
   61 shards
   TQ groups: 47616, regular: 1123
@@ -231,7 +231,7 @@ quantized MiniMax on Apple Silicon.
 JANGTQ takes this one step further by using a learned codebook for the 2-bit
 expert weights. For MiniMax M2.5, JANG_2L (affine) scored 74% MMLU vs MLX's
-25%. For MiniMax M2.7, **JANGTQ_2L scores 91.5%** — the highest-quality
 sub-60-GB MiniMax quant on any runtime.
 ---

 <div align="center">
+# MiniMax-M2.7 JANGTQ
 **MiniMax M2.7 228B MoE — 2.15-bit codebook + Hadamard, 56.5 GB**
 Result: **smaller than affine 2-bit, higher quality than affine 2-bit, runs at
 89% of affine 2-bit speed** on Apple Silicon.
+| | JANG_2L (affine) | **JANGTQ** | Δ |
 |---|---|---|---|
 | Disk size | ~63 GB | **56.5 GB** | **−10%** |
 | GPU memory | ~62.6 GB | **56.5 GB** | **−10%** |
 Tested 2026-04-13 on Mac Studio M3 Ultra. Reasoning enabled (MiniMax M2.7 is
 an always-reasoning model); `<think>…</think>` stripped before scoring.
+| Subject | JANGTQ | JANG_2L (affine) | JANG_3L/4M |
 |---|---|---|---|
 | **astronomy** | **20/20 (100%)** | — | — |
 | **high_school_biology** | **20/20 (100%)** | — | — |
 | logical_fallacies | 16/20 (80%) | — | — |
 | **Total** | **183/200 = 91.5%** | **~88%** | **~95.5%** |
+JANGTQ sits cleanly between affine JANG_2L (88%) and the larger JANG_3L/4M
 (95.5%) — capturing most of the quality of the 3L/4M profiles at ~55-60% of
 their disk footprint.
 | Architecture | MoE (256 experts, top-8 active), standard Q/K/V attention, partial RoPE |
 | Total parameters | 228.7 B |
 | Active per token | ~1.4 B |
+| Profile | **JANGTQ** |
 | Format | **JANGTQ (codebook+Hadamard)** — `weight_format: mxtq` in `jang_config.json` |
 | Avg bits/param | ~2.15 |
 | Disk | **56.55 GB** |
 | Context | 192 K tokens |
 | Chat template | Always-reasoning (`<think>\n` opened at assistant start) |
+## JANGTQ Bit Allocation
 | Component | Bits | Format | Why |
 |---|---|---|---|
 from jang_tools.load_jangtq import load_jangtq_model
 from mlx_lm import generate
+model_path = snapshot_download("JANGQ-AI/MiniMax-M2.7-JANGTQ")
 model, tokenizer = load_jangtq_model(model_path)
 messages = [{"role": "user", "content": "Explain photosynthesis in 5 sentences."}]
 On first load you'll see log lines like:
 ```
+Loading JANGTQ: MiniMax-M2.7-JANGTQ
   seed=42, bits_map={'attention': 8, ..., 'routed_expert': 2, ...}
   61 shards
   TQ groups: 47616, regular: 1123
 JANGTQ takes this one step further by using a learned codebook for the 2-bit
 expert weights. For MiniMax M2.5, JANG_2L (affine) scored 74% MMLU vs MLX's
+25%. For MiniMax M2.7, **JANGTQ scores 91.5%** — the highest-quality
 sub-60-GB MiniMax quant on any runtime.
 ---