Instructions to use ManniX-ITA/Qwen3.5-4B-M2-OMv2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use ManniX-ITA/Qwen3.5-4B-M2-OMv2 with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf ManniX-ITA/Qwen3.5-4B-M2-OMv2:Q6_K
# Run inference directly in the terminal:
llama cli -hf ManniX-ITA/Qwen3.5-4B-M2-OMv2:Q6_K

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf ManniX-ITA/Qwen3.5-4B-M2-OMv2:Q6_K
# Run inference directly in the terminal:
llama cli -hf ManniX-ITA/Qwen3.5-4B-M2-OMv2:Q6_K

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 ManniX-ITA/Qwen3.5-4B-M2-OMv2:Q6_K
# Run inference directly in the terminal:
./llama-cli -hf ManniX-ITA/Qwen3.5-4B-M2-OMv2:Q6_K

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 ManniX-ITA/Qwen3.5-4B-M2-OMv2:Q6_K
# Run inference directly in the terminal:
./build/bin/llama-cli -hf ManniX-ITA/Qwen3.5-4B-M2-OMv2:Q6_K

Use Docker

docker model run hf.co/ManniX-ITA/Qwen3.5-4B-M2-OMv2:Q6_K

LM Studio
Jan

vLLM

How to use ManniX-ITA/Qwen3.5-4B-M2-OMv2 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ManniX-ITA/Qwen3.5-4B-M2-OMv2"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ManniX-ITA/Qwen3.5-4B-M2-OMv2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/ManniX-ITA/Qwen3.5-4B-M2-OMv2:Q6_K

Ollama
How to use ManniX-ITA/Qwen3.5-4B-M2-OMv2 with Ollama:
```
ollama run hf.co/ManniX-ITA/Qwen3.5-4B-M2-OMv2:Q6_K
```
Unsloth Desktop

How to use ManniX-ITA/Qwen3.5-4B-M2-OMv2 with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf ManniX-ITA/Qwen3.5-4B-M2-OMv2:Q6_K

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "ManniX-ITA/Qwen3.5-4B-M2-OMv2:Q6_K"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use ManniX-ITA/Qwen3.5-4B-M2-OMv2 with Docker Model Runner:
```
docker model run hf.co/ManniX-ITA/Qwen3.5-4B-M2-OMv2:Q6_K
```

Lemonade

How to use ManniX-ITA/Qwen3.5-4B-M2-OMv2 with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull ManniX-ITA/Qwen3.5-4B-M2-OMv2:Q6_K

Run and chat with the model

lemonade run user.Qwen3.5-4B-M2-OMv2-Q6_K

List all available models

lemonade list

Hermes Agent

How to use ManniX-ITA/Qwen3.5-4B-M2-OMv2 with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf ManniX-ITA/Qwen3.5-4B-M2-OMv2:Q6_K

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 ManniX-ITA/Qwen3.5-4B-M2-OMv2:Q6_K

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use ManniX-ITA/Qwen3.5-4B-M2-OMv2 with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf ManniX-ITA/Qwen3.5-4B-M2-OMv2:Q6_K

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 "ManniX-ITA/Qwen3.5-4B-M2-OMv2:Q6_K" \
  --custom-provider-id llama-cpp \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

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

Qwen3.5-4B-M2-OMv2

OMv2 recipe (OBIM-lite + DAREx-q + EMR election) without any importance-signal weighting. Isolates the recipe's contribution from the importance signal in the M3/M5 variants.

Sources

Role	Model
Base	`Qwen/Qwen3.5-4B`
Source A	`Jackrong/Qwen3.5-4B-Claude-4.6-Opus-Reasoning-Distilled-v2`
Source B	`Crownelius/Crow-4B-Opus-4.6-Distill-Heretic_Qwen3.5`

Weights: 0.55 (A) / 0.45 (B). Density: 0.53. Seed: 42.

Phase 1+2 comparison (Q6_K)

#	Recipe	Merger	Importance	HumanEval pass@1	MBPP pass@1
floor	Qwen/Qwen3.5-4B (no merge)	—	—	60.37% 🥇	45.80%
src A	Jackrong-v2 (Claude-4.6-Opus reasoning distill)	—	—	60.37%	45.00%
src B	Crow-4B (Opus-4.6 distill, heretic)	—	—	54.88%	48.20%
M1	Vanilla DARE-TIES	dare_ties_merge.py	none	51.22%	47.00%
M2 (this)	OMv2 recipe (OBIM-lite + DAREx-q + EMR election)	dare_ties_merge.py	none	52.44%	49.40%
M3	OMv2 + Fisher	dare_ties_merge.py	Fisher	57.93%	48.80%
M4	ex-LRP (mergekit PR #682)	mergekit (PR #682)	LRP	51.22%	49.40%
M4-v2	ex-LRP (PR #682 turbo, w=1/1, d=0.7)	mergekit PR #682 turbo	LRP	55.49%	52.20% 🥇
M5	OMv2 + LRP	dare_ties_merge.py	LRP	53.05%	51.40%

Eval methodology: llama-server (--reasoning-format deepseek --reasoning-budget 8192 --parallel 2 --cache-type-k q8_0 --cache-type-v q8_0 -c 32768) → lm_eval local-completions against raw /v1/completions, temperature 0, max_gen_toks=2048. All five variants scored under identical conditions. Floor reference: the top three rows are the base model and the two source fine-tunes evaluated under the same methodology. Hard finding: no merge beats Qwen3.5-4B base on HumanEval (best merge M3 = 57.93%, base = 60.37%). On MBPP, M4-v2 (52.20%) and M5 (51.40%) clearly beat both sources (best source: Crow at 48.20%) — merging does add MBPP capability, but borrows HumanEval ground.