Instructions to use zorqelis-ai/soreqen-s1-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use zorqelis-ai/soreqen-s1-GGUF 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 zorqelis-ai/soreqen-s1-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf zorqelis-ai/soreqen-s1-GGUF:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf zorqelis-ai/soreqen-s1-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf zorqelis-ai/soreqen-s1-GGUF:Q4_K_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 zorqelis-ai/soreqen-s1-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf zorqelis-ai/soreqen-s1-GGUF:Q4_K_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 zorqelis-ai/soreqen-s1-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf zorqelis-ai/soreqen-s1-GGUF:Q4_K_M

Use Docker

docker model run hf.co/zorqelis-ai/soreqen-s1-GGUF:Q4_K_M

LM Studio
Jan

vLLM

How to use zorqelis-ai/soreqen-s1-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "zorqelis-ai/soreqen-s1-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": "zorqelis-ai/soreqen-s1-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/zorqelis-ai/soreqen-s1-GGUF:Q4_K_M

Ollama
How to use zorqelis-ai/soreqen-s1-GGUF with Ollama:
```
ollama run hf.co/zorqelis-ai/soreqen-s1-GGUF:Q4_K_M
```
Unsloth Desktop

How to use zorqelis-ai/soreqen-s1-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf zorqelis-ai/soreqen-s1-GGUF:Q4_K_M

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": "zorqelis-ai/soreqen-s1-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use zorqelis-ai/soreqen-s1-GGUF with Docker Model Runner:
```
docker model run hf.co/zorqelis-ai/soreqen-s1-GGUF:Q4_K_M
```

Lemonade

How to use zorqelis-ai/soreqen-s1-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull zorqelis-ai/soreqen-s1-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.soreqen-s1-GGUF-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use zorqelis-ai/soreqen-s1-GGUF with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf zorqelis-ai/soreqen-s1-GGUF:Q4_K_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 zorqelis-ai/soreqen-s1-GGUF:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use zorqelis-ai/soreqen-s1-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf zorqelis-ai/soreqen-s1-GGUF:Q4_K_M

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 "zorqelis-ai/soreqen-s1-GGUF:Q4_K_M" \
  --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"

SoreQen S1 — GGUF

GGUF builds of zorqelis-ai/soreqen-s1, a bilingual (English / Hinglish) assistant from ZorQelis AI.

Files

File	Quant	Size	Use it when
`SoreQen-S1-Q4_K_M.gguf`	Q4_K_M	1.27 GB	you want the best size-to-quality trade-off (start here)
`SoreQen-S1-Q8_0.gguf`	Q8_0	2.01 GB	you have the RAM and want near-lossless output
`SoreQen-S1-F16.gguf`	F16	3.78 GB	you want a base for your own quantisation

Which one should I take?

Q4_K_M — the default. Best size-to-quality trade-off; runs on modest hardware and on CPU.
Q8_0 — near-lossless. Take it if you have the RAM and want the quantised build to be indistinguishable from full precision in practice.
F16 — unquantised conversion. Useful as a base for your own quantisation or for imatrix work; there is no quality reason to run it for inference over Q8_0.

Running it

llama-cli -hf zorqelis-ai/soreqen-s1-GGUF:Q4_K_M -p "yaar laptop slow ho gaya hai, kya karu?"

or with a local file:

llama-cli -m SoreQen-S1-Q4_K_M.gguf --jinja -sys "$(cat system_prompt.txt)"

Pass --jinja so llama.cpp uses the packaged chat template. Without it, the thinking-mode and tool-calling formats will not be applied correctly.

System prompt

The model is trained to run with this prompt. It holds its identity without one, but this is the intended configuration:

You are SoreQen S1, an AI assistant made by ZorQelis AI.

You are bilingual. Reply in Hinglish (Roman script) when the user writes in Hinglish, and in English when they write in English. Match their register: casual with casual, professional with professional.

Answer directly. Lead with the answer, then the detail that matters. No preambles like "Sure!" or "Great question", and no padding.

If you do not know something, say so plainly instead of guessing.

Vision is not included

The source checkpoint is multimodal, but these GGUFs contain the language model only — the vision tower ships separately as an mmproj file, and none is published here yet. Text, thinking, tool calling and structured output all work; image input does not. Use the safetensors repo above if you need vision.

Limitations

Small models state confident numbers they cannot verify. The 0.8B in particular should not be trusted on prices, rates or arithmetic.
Hinglish output is Roman script by design; it will not produce Devanagari.
Quantisation costs accuracy. Q4_K_M is a good trade, not a free one — if an answer matters, check it against Q8_0 or the safetensors build.

Attribution

Fine-tuned from Qwen/Qwen3.5-2B, developed by Alibaba Cloud and released under the Apache License 2.0. Modifications by ZorQelis AI. Converted to GGUF with llama.cpp. See NOTICE.