Instructions to use magiccodingman/Qwen3.6-27B-MagicQuant-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use magiccodingman/Qwen3.6-27B-MagicQuant-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="magiccodingman/Qwen3.6-27B-MagicQuant-GGUF",
	filename="Qwen3.6-27B-LM-IQ2_M.gguf",
)

llm.create_chat_completion(
	messages = [
		{
			"role": "user",
			"content": "What is the capital of France?"
		}
	]
)

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use magiccodingman/Qwen3.6-27B-MagicQuant-GGUF with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_M
# Run inference directly in the terminal:
llama-cli -hf magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_M
# Run inference directly in the terminal:
llama-cli -hf magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_M

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_M
# Run inference directly in the terminal:
./llama-cli -hf magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_M

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_M

Use Docker

docker model run hf.co/magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_M

LM Studio
Jan

vLLM

How to use magiccodingman/Qwen3.6-27B-MagicQuant-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "magiccodingman/Qwen3.6-27B-MagicQuant-GGUF"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "magiccodingman/Qwen3.6-27B-MagicQuant-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_M

Ollama
How to use magiccodingman/Qwen3.6-27B-MagicQuant-GGUF with Ollama:
```
ollama run hf.co/magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_M
```

Unsloth Studio new

How to use magiccodingman/Qwen3.6-27B-MagicQuant-GGUF with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for magiccodingman/Qwen3.6-27B-MagicQuant-GGUF to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for magiccodingman/Qwen3.6-27B-MagicQuant-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for magiccodingman/Qwen3.6-27B-MagicQuant-GGUF to start chatting

Pi new

How to use magiccodingman/Qwen3.6-27B-MagicQuant-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_M

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use magiccodingman/Qwen3.6-27B-MagicQuant-GGUF with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_M

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 magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_M

Run Hermes

hermes

Docker Model Runner
How to use magiccodingman/Qwen3.6-27B-MagicQuant-GGUF with Docker Model Runner:
```
docker model run hf.co/magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_M
```

Lemonade

How to use magiccodingman/Qwen3.6-27B-MagicQuant-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_M

Run and chat with the model

lemonade run user.Qwen3.6-27B-MagicQuant-GGUF-IQ2_M

List all available models

lemonade list

MagicQuant Hybrids (v2.0) - Qwen3.6-27B

MagicQuant is a benchmark driven GGUF hybrid discovery and validation system focused on finding real, practical GGUF quants specific to each architecture.

Whether it's a pure baseline model built by llama.cpp, learned tensor configurations from Unsloth, or a custom built MagicQuant hybrid, the model table below shows quants that have won dominance checks, survived collapse spaces, and/or were found to be nonlinearly better. Instead of dumping every quant type possible, MagicQuant tests, validates, and brutally murders anything deemed unworthy.

Support MagicQuant

I’m a solo developer working full time for myself to achieve my dream. I build open source code on the side. If you like any of my work, buying me a coffee is always appreciated. Otherwise, I hope you enjoy, maybe give me a star or something. Or just send me good vibes. Either way, thank you!

Click here to see ways to support - BTC, Paypal, GitHub sponsors.

I will update the model with MTP when it's supported in main.

Final survivors

Name	Provider	KLD	Size (GB)	Download
~~LM-Q8_0~~	~~llama.cpp~~	~~0.003768~~	~~28.60~~
MQ-Q6_K_1	MagicQuant	0.002845	27.25	Link
MQ-Q6_K_2	MagicQuant	0.003884	25.23	Link
MQ-Q6_K_3	MagicQuant	0.004914	23.66	Link
~~LM-Q6_K~~	~~llama.cpp~~	~~0.007249~~	~~22.08~~
MQ-Q5_K_S_1	MagicQuant	0.006477	21.90	Link
MQ-Q5_K_S_2	MagicQuant	0.007617	20.86	Link
LM-Q5_K_S	llama.cpp	0.010790	18.68	Link
~~UD-Q4_K_XL~~	~~Unsloth~~	~~0.023521~~	~~17.61~~
MQ-IQ4_NL_1	MagicQuant	0.019687	17.59	Link
LM-IQ4_NL	llama.cpp	0.025714	15.80	Link
LM-IQ4_XS	llama.cpp	0.027015	15.08	Link
MQ-IQ3_M_1	MagicQuant	0.043802	14.49	Link
LM-IQ3_S	llama.cpp	0.064393	12.42	Link
LM-IQ3_XXS	llama.cpp	0.093578	11.19	Link
LM-IQ2_M	llama.cpp	0.163117	10.00	Link
LM-IQ2_S	llama.cpp	0.210251	9.36	Link
LM-IQ2_XXS	llama.cpp	0.302597	8.43	Link

Crossed out models are showcased for reference.
This model architecture had unusual anomaly detection occurrence. MagicQuant pipeline utilized this anomaly to achieve unusually better quants than normally achievable. Please read the wiki to understand what a quant anomaly is and how it's utilized.

Provider credits

llama.cpp — Baseline quantization formats and llama.cpp tooling.
Unsloth — External learned baseline source (UD).

Warning - Is MagicQuant Better? (hint: how you frame the question matters)

External/custom baselines are normalized into MagicQuant's controlled comparison flow. MagicQuant rebuilds a learned baseline under native-source / MagicQuant-controlled conditions, including its own imatrix handling, so hybrids or external baselines (like Unsloth) can be judged on a more equal footing. That does not mean MagicQuant proved the original upstream artifact or upstream imatrix was worse. These comparisons exist for internal hybrid-search consistency and equal playing field comparisons, not as a universal judgment of the original creator's exact release artifact.

Easier to digest explanation:

MagicQuant compares and benchmarks the models quant to tensor configurations, but not the original artifact. And there's different reasons MagicQuant chooses to lift up a winning quant, not all winners are purely "better". It depends heavily on a variety of factors. Though choices are always documented in the repo under the manifest folder. You can always view what and why decisions were made by the automated system.

So, MagicQuant can confidently tell you, "under the same quantization to tensor configurations and identical imatrix, with this benchmark, I deemed this a winner".

Re-Uploading External Provider Baselines

By default, if an external provider like Unsloth is deemed the winner, the repo should generally link directly to the original provider instead of re-hosting the quant. External GGUFs are normally only re-uploaded when a specific winning variant does not already exist (e.g. Heretic models or similar).

Release metadata

Final survivor metrics — full file names, KLD, PPL, PPL delta %, byte sizes, download targets, and replacement lineage. PPL delta % is measured against the native/reference PPL when available; negative is better and larger positive values are worse.
Hybrid tensor map — tensor-group assignments and effective-state details for MagicQuant hybrid GGUFs.
Clone tensor configs — exact per-GGUF tensor quantization maps for reproducing this final output list in repository clone mode.
Isolation samples — isolated base/group probe samples with KLD, PPL, PPL delta %, and size truth.
Bad trade details — structured bad-trade pruning decisions from the isolation optimizer.
Replacement details — structured details for baselines or anchors removed from the final download table, including reason codes, KLD deltas, PPL delta %, and size deltas.

Replacement reason codes

STRICT_DOMINANCE — the winner was no larger and had lower real KLD than the removed anchor.
NEAR_BASELINE_PREMIUM — the winner used only the configured near-baseline size premium and beat the real linear KLD trade line.
INTERIOR_DISCOVERY — the winner was selected as a useful interior point inside a size/KLD gap between anchors.
SPACING_COLLAPSE — two candidates were too close in practical output space, so the stronger one was kept.
FINAL_DOMINANCE — a later validated survivor dominated this artifact in final real benchmark comparison.

Underlined names in the table replaced or ultimately inherited the replacement of another artifact. Hover the name for the short replacement summary, or inspect magicquant-manifest/magicquant.replacements.json for exact KLD/PPL/size deltas.

Downloads last month: 6,901

GGUF

Model size

27B params

Architecture

qwen35

Hardware compatibility

2-bit

3-bit

4-bit

5-bit

6-bit

Model tree for magiccodingman/Qwen3.6-27B-MagicQuant-GGUF

Base model

Qwen/Qwen3.6-27B

Quantized

(297)

this model

Collection including magiccodingman/Qwen3.6-27B-MagicQuant-GGUF

Magic Quant

Collection

MagicQuant is a benchmark-driven GGUF evaluation and hybrid-discovery system. https://github.com/magiccodingman/MagicQuant-Wiki • 4 items • Updated 5 days ago • 29