Instructions to use Parakon/Parakon-122B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use Parakon/Parakon-122B 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 Parakon/Parakon-122B
# Run inference directly in the terminal:
llama cli -hf Parakon/Parakon-122B

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf Parakon/Parakon-122B
# Run inference directly in the terminal:
llama cli -hf Parakon/Parakon-122B

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 Parakon/Parakon-122B
# Run inference directly in the terminal:
./llama-cli -hf Parakon/Parakon-122B

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 Parakon/Parakon-122B
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Parakon/Parakon-122B

Use Docker

docker model run hf.co/Parakon/Parakon-122B

LM Studio
Jan

vLLM

How to use Parakon/Parakon-122B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Parakon/Parakon-122B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Parakon/Parakon-122B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Parakon/Parakon-122B

Ollama
How to use Parakon/Parakon-122B with Ollama:
```
ollama run hf.co/Parakon/Parakon-122B
```
Unsloth Desktop

How to use Parakon/Parakon-122B with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Parakon/Parakon-122B

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": "Parakon/Parakon-122B"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use Parakon/Parakon-122B with Docker Model Runner:
```
docker model run hf.co/Parakon/Parakon-122B
```

Lemonade

How to use Parakon/Parakon-122B with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Parakon/Parakon-122B

Run and chat with the model

lemonade run user.Parakon-122B-{{QUANT_TAG}}

List all available models

lemonade list

Hermes Agent

How to use Parakon/Parakon-122B with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Parakon/Parakon-122B

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 Parakon/Parakon-122B

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use Parakon/Parakon-122B with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Parakon/Parakon-122B

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 "Parakon/Parakon-122B" \
  --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"

You need to agree to share your contact information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Parakon 122B

Qwen3.5-122B-A10B compressed by Parakon's pipeline: 244 GB (bf16) → 23.87 GB, a 10.2× reduction. To our knowledge the smallest published artifact of this model, by roughly 30%.

Quality — in progress, published as it lands

Scores below are the compressed model's results as a share of the model's published full-precision reference scores. The full suite is still running; benchmarks are published as they complete rather than held back until the average flatters us.

Benchmark	Retention vs published reference
MMLU-Redux (knowledge)	93.5%
IFEval (instruction following)	93.6%
GPQA-Diamond (hard science)	69.4%
Average, three complete	85.5%
Remaining benchmarks	still running

Knowledge and instruction following retain better than nine tenths of the reference. Long-chain scientific reasoning is where the compression costs, and it is reported at the same size as the rest.

Serving — measured


CPU-only inference	supported
Round-trip export check	bit-exact

Runtime

This artifact uses Parakon's own storage format and requires the Parakon runtime (custom GPU and CPU kernels) to execute. The runtime is provided to evaluation partners together with reproduction instructions for every number above.

License & access

Released under the Parakon Community License:

Always free — research, personal use, evaluation, and benchmarking (publishing your results is encouraged, never restricted)
Free commercial use for organizations under 100 employees and $1M annual revenue — production included
Larger organizations need a commercial agreement — contact the Parakon team through this organization's page
No re-hosting — refer others to this repository for the weights

The runtime is licensed separately. For deployment licensing, runtime access, or compression engagements on your own models: get in touch.

Downloads last month: -

GGUF

Model size

122B params

Architecture

qwen35moe

Hardware compatibility

We're not able to determine the quantization variants.

View all variants

Model tree for Parakon/Parakon-122B

Base model

Qwen/Qwen3.5-122B-A10B

Quantized

(155)

this model