Instructions to use Solel-io/Angel-1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use Solel-io/Angel-1 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 Solel-io/Angel-1:IQ4_XS
# Run inference directly in the terminal:
llama cli -hf Solel-io/Angel-1:IQ4_XS

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf Solel-io/Angel-1:IQ4_XS
# Run inference directly in the terminal:
llama cli -hf Solel-io/Angel-1:IQ4_XS

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 Solel-io/Angel-1:IQ4_XS
# Run inference directly in the terminal:
./llama-cli -hf Solel-io/Angel-1:IQ4_XS

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 Solel-io/Angel-1:IQ4_XS
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Solel-io/Angel-1:IQ4_XS

Use Docker

docker model run hf.co/Solel-io/Angel-1:IQ4_XS

Llama new
LM Studio
Jan

vLLM

How to use Solel-io/Angel-1 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Solel-io/Angel-1"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Solel-io/Angel-1",
		"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/Solel-io/Angel-1:IQ4_XS

Ollama
How to use Solel-io/Angel-1 with Ollama:
```
ollama run hf.co/Solel-io/Angel-1:IQ4_XS
```
Unsloth Desktop

How to use Solel-io/Angel-1 with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Solel-io/Angel-1:IQ4_XS

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": "Solel-io/Angel-1:IQ4_XS"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use Solel-io/Angel-1 with Docker Model Runner:
```
docker model run hf.co/Solel-io/Angel-1:IQ4_XS
```

Lemonade

How to use Solel-io/Angel-1 with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Solel-io/Angel-1:IQ4_XS

Run and chat with the model

lemonade run user.Angel-1-IQ4_XS

List all available models

lemonade list

Hermes Agent

How to use Solel-io/Angel-1 with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Solel-io/Angel-1:IQ4_XS

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 Solel-io/Angel-1:IQ4_XS

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use Solel-io/Angel-1 with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Solel-io/Angel-1:IQ4_XS

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 "Solel-io/Angel-1:IQ4_XS" \
  --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"

Aether Series by Solel//ai

Angel-1

Angel is the fast one. It's the smallest model in the Aether Series from Solel//ai, and the default model in Solel//os. It's built to run on almost anything you've got at home: a modest graphics card, a mini PC, or a decent laptop.

Angel reads images as well as text. Give it a photo, a screenshot or a scanned page and ask about it.

All sunshine. No cloud.

At a glance


Family	Aether Series by Solel//ai
Role	Fast, everyday default
Parameters	About 4.7 billion (language model plus vision encoder)
Built on	Qwen3.5-4B (Apache-2.0)
Input	Text and images
Output	Text
Context window	Up to 262,144 tokens (in practice, set by your memory)
Format	GGUF, IQ4_XS quantisation, with an f16 vision projector
Licence	Apache-2.0, open source

Files

File	Size	What it is
`Angel-1.IQ4_XS.gguf`	2.45 GB	The model weights
`Angel-1.mmproj-f16.gguf`	0.67 GB	The vision projector, needed for images
`chat_template.jinja`, `tokenizer.json`, `config.json` and friends	small	Template, tokeniser and config

Hardware

About 3.1 GB of files in total. A graphics card with 4 GB of memory will run it with a short context; 6 GB or more gives you room for longer chats. It also runs on a CPU, just slower.

Run it

In Solel//os

Angel comes installed. Nothing to do.

With llama.cpp

llama-server -m Angel-1.IQ4_XS.gguf --mmproj Angel-1.mmproj-f16.gguf

Then point any OpenAI-compatible client at the server and attach images to your messages.

Good at, and not so good at

Angel is quick and light, which means trade-offs. It's good for chat, summaries, describing images, reading text in screenshots and everyday questions. It's weaker than Dominion or Seraphim on long reasoning, maths and detailed knowledge, and like every language model it can be confidently wrong. Check anything that matters.

Content and safety

Angel has lighter refusal behaviour than its base model. It's meant for adults running it on their own hardware. If you put it in front of other people, add your own filtering. You're responsible for how you use it and what you do with its output.

Licence

Angel-1 is released under the Apache License 2.0. It's built on Qwen3.5-4B by the Qwen team, also under Apache-2.0; their licence and notices are kept with these files.

The names Solel, Solel//ai, Aether Series and Angel are trade marks of Solel//io. The licence covers the weights, not the names. See the Solel//io trade mark policy.

About Solel//ai

solel//ai

Solel//ai is the AI division of Solel//io, an Australian open-source home server business. We make local models that run on hardware you own, so your conversations stay in your house.