Instructions to use MarsupialAI/SkunkApe-14b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use MarsupialAI/SkunkApe-14b with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="MarsupialAI/SkunkApe-14b")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("MarsupialAI/SkunkApe-14b")
model = AutoModelForCausalLM.from_pretrained("MarsupialAI/SkunkApe-14b")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use MarsupialAI/SkunkApe-14b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "MarsupialAI/SkunkApe-14b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "MarsupialAI/SkunkApe-14b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/MarsupialAI/SkunkApe-14b

SGLang

How to use MarsupialAI/SkunkApe-14b with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "MarsupialAI/SkunkApe-14b" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "MarsupialAI/SkunkApe-14b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "MarsupialAI/SkunkApe-14b" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "MarsupialAI/SkunkApe-14b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use MarsupialAI/SkunkApe-14b with Docker Model Runner:
```
docker model run hf.co/MarsupialAI/SkunkApe-14b
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

Skunk Ape 14b

This version performs substantially better than the 16b version.

This model is a merge of three self-merged Solar-based models in a 14b (64 layer) configuration. The result of this "frankenmerge" is a medium-sized model that contains what I consider to be the best of the solar finetunes.

Mergefuel:

Sao10K/Fimbulvetr-11B-v2
Sao10K/Solstice-11B-v1
TheDrummer/Moistral-11B-v1

This model is uncensored and capable of generating objectionable material. However, it is not an explicitely-NSFW model, and it has never "gone rogue" and tried to insert NSFW content into SFW prompts in my experience. As with any LLM, no factual claims made by the model should be taken at face value. You know that boilerplate safety disclaimer that most professional models have? Assume this has it too. This model is for entertainment purposes only.

iMatrix GGUFs: https://huggingface.co/MarsupialAI/SkunkApe-14b_iMatrix_GGUF

Sample output

{{[INPUT]}}
Write a detailed and humorous story about a cute and fluffy bunny that goes to a Gwar concert.
{{[OUTPUT]}}
Bobby was the softest, fluffiest, and cutest rabbit you had ever seen. His thick fur, bright eyes, and cuddly little face would make anyone who laid their eyes upon him just want to squeeze him and make him their best friend forever. But there was something no one knew about Bobby: he was a heavy metalhead at heart. 

That Saturday morning, Bobby hopped out of his pen in his owners' backyard, excited for what the day had in store for him. Today, he was going on an adventure. Not just any adventure, but an epic heavy metal adventure. Today, he was going to see the mighty Gwar in concert. Ever since he first heard their music, Bobby knew he needed to experience their raw power live.

He hopped through the neighborhood and made his way towards town. He was so excited; his heart was beating like a double kick drum. Eventually, he got to the venue and found himself standing in line with a group of leather-clad headbangers. They took one look at this innocent looking bunny with a denim jacket, black band patches, and a Slayer baseball cap and began to laugh. Bobby didn’t mind; it just fueled his fire to prove these posers wrong.

A burly bouncer came up to Bobby as he approached the front of the line, "Um, buddy. No rabbits allowed."

<<<This goes on for a while.  See sample.txt for full output>>>

Prompt format

Prefers alpaca.

Weird merge fuckery

According to Toasty Pigeon, FM, Akai, and probably others on the KAI discord, this merge method works better than a normal stacked merge.
I don't pretend to understand why, but the huge PPL improvement (5.96 for this model vs 7.65 for the 16b @ Q4km) indicates that they're right.
See recipe.txt for all the alchemy.

Downloads last month: 6

Safetensors

Model size

14B params

Tensor type

F16

Model tree for MarsupialAI/SkunkApe-14b

Finetunes

1 model

Quantizations

2 models

Spaces using MarsupialAI/SkunkApe-14b 15

Collection including MarsupialAI/SkunkApe-14b

Merges

Collection

Stuff I've cooked myself, for better or for worse. • 14 items • Updated Jun 18, 2025 • 3