Instructions to use gbuzhf/SNEX-V1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use gbuzhf/SNEX-V1 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 gbuzhf/SNEX-V1:BF16
# Run inference directly in the terminal:
llama cli -hf gbuzhf/SNEX-V1:BF16

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf gbuzhf/SNEX-V1:BF16
# Run inference directly in the terminal:
llama cli -hf gbuzhf/SNEX-V1:BF16

Build from source code

git clone https://github.com/ggml-org/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-app
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama serve -hf gbuzhf/SNEX-V1:BF16
# Run inference directly in the terminal:
./build/bin/llama cli -hf gbuzhf/SNEX-V1:BF16

Use Docker

docker model run hf.co/gbuzhf/SNEX-V1:BF16

Llama new
LM Studio
Jan

vLLM

How to use gbuzhf/SNEX-V1 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "gbuzhf/SNEX-V1"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "gbuzhf/SNEX-V1",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/gbuzhf/SNEX-V1:BF16

Ollama
How to use gbuzhf/SNEX-V1 with Ollama:
```
ollama run hf.co/gbuzhf/SNEX-V1:BF16
```
Unsloth Desktop

How to use gbuzhf/SNEX-V1 with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf gbuzhf/SNEX-V1:BF16

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": "gbuzhf/SNEX-V1:BF16"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use gbuzhf/SNEX-V1 with Docker Model Runner:
```
docker model run hf.co/gbuzhf/SNEX-V1:BF16
```

Lemonade

How to use gbuzhf/SNEX-V1 with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull gbuzhf/SNEX-V1:BF16

Run and chat with the model

lemonade run user.SNEX-V1-BF16

List all available models

lemonade list

Hermes Agent

How to use gbuzhf/SNEX-V1 with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf gbuzhf/SNEX-V1:BF16

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 gbuzhf/SNEX-V1:BF16

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use gbuzhf/SNEX-V1 with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf gbuzhf/SNEX-V1:BF16

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 "gbuzhf/SNEX-V1:BF16" \
  --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"

SNEX-V1

pinned

by gbuzhf - opened 7 days ago

Discussion

gbuzhf

Owner 7 days ago

•

edited 6 days ago

Experimental merge: Sage-A1.5 (a Qwen3.6-35B-A3B fine-tune) is the base, plus a small slice of Nex-N2.5-mini's own training.
From every tensor, the top 25% of Nex's changes against its own base (Qwen3.5-35B-A3B), about 75% of that change's energy, are added at a quarter strength. Sage's router, norms, embeddings and MTP head stay frozen.
Normalization -> refitting -> BF16 -> imatrix -> ICE 1.6 -> repair/refitting while imatrix activated (ICE 2.0) -> specialized Arrow4.0 -> measurements -> Release.

Short test (Oct 24, 2026): SNEX-V1-25G-ICEv2 vs official Sage-A1.5-Q6_K on the same 19 Terminal-Bench 2.1 tasks, Hermes Agent, identical GPUs, 3× time limits, one run each:

	SNEX-V1	Sage-A1.5
Tasks solved	7 / 19	12 / 19
Tests passed	61.8 %	77.2 %

On these tasks the transplant did not help; Sage scored higher. A small sample, not a full benchmark.

gbuzhf pinned discussion 7 days ago

gbuzhf

Owner 7 days ago

Vision Tower (Q8)

News-Curation Task

gbuzhf unpinned discussion 7 days ago

gbuzhf pinned discussion 7 days ago

gbuzhf

Owner 7 days ago

•

edited 6 days ago

Heavy benchmarks are running. I will publish on model card with raw results uploaded.
//edit: So nex's delta energy didn't make this more intelligent, instead, seemingly on Terminal Bench 2.1 (especially the one I targeted) it regressed.

I will prod ICE 2.0 for next Run.

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment