Instructions to use Kwaipilot/KAT-Dev with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Kwaipilot/KAT-Dev with Transformers:

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

pipe = pipeline("text-generation", model="Kwaipilot/KAT-Dev")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Kwaipilot/KAT-Dev")
model = AutoModelForCausalLM.from_pretrained("Kwaipilot/KAT-Dev")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Inference
Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use Kwaipilot/KAT-Dev with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/Kwaipilot/KAT-Dev

SGLang

How to use Kwaipilot/KAT-Dev 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 "Kwaipilot/KAT-Dev" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Kwaipilot/KAT-Dev",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

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 "Kwaipilot/KAT-Dev" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Kwaipilot/KAT-Dev",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Kwaipilot/KAT-Dev with Docker Model Runner:
```
docker model run hf.co/Kwaipilot/KAT-Dev
```

Is KAT-Dev supposed to be that much slower than QWEN?

#11

by deleted - opened Oct 17, 2025

Discussion

deleted

Oct 17, 2025

•

edited Oct 17, 2025

Hello, I tried both KAT-Dev (33B) and QWEN3-Coder-30B-A3B to compare speed. QWEN is significantly faster. I don't know about quality though, but I assume KAT-Dev is better from hearing from other people.

My stats:

KAT-Dev GGUF Q5_K_M: 16 tk/s
KAT-Dev GGUF Q4_K_M: 27 tk/s
QWEN3-Coder GGUF Q5_K_XL: 201 tk/s

Load settings:

Both KAT-Dev GGUFs: Just V Cache Quantization Type set to Q8_0 (with Flash Attention on), rest default
QWEN3-Coder GGUF: K- and V Cache Quantization Type set to Q8_0 (with Flash Attention on), rest default

Inference settings (favored KAT-Dev settings):

All 3 with same inference settings, which is:
Temp: 0.6,
Top K: 20,
Repeat Penatly: 1.05,
Min P Sampling: 0,
Top P Sampling: 0.95

1 x RTX 5090 (32 GB total VRAM)
2 x 32 GB RAM (64 GB total RAM) @ 5600 MHz
1 x Ryzen 9 7950x

YoLo2000

Kwaipilot org Oct 17, 2025

Thanks for the benchmark! The speed difference is expected - KAT-Dev is a dense 32B model that processes all parameters per token, while QWEN3-Coder is MoE with only ~3B activated parameters.
Your results (7.4x faster for QWEN) match the theory pretty well, since 32B/3B ≈ 10x. Dense models like KAT-Dev are more thorough but slower, while MoE gets you speed through selective activation.
Appreciate the detailed comparison!

deleted

Oct 17, 2025

Thanks, good to hear

Reverger

Oct 17, 2025

@sk0d
If Kwaipilot releases 0.6b or 14b model, you can use so-called "draft model" acceleration like this:
https://developer.nvidia.com/blog/boost-llama-3-3-70b-inference-throughput-3x-with-nvidia-tensorrt-llm-speculative-decoding/

akierum

Oct 22, 2025

Do you use LMstudio? does it work with roocode/cline or you?

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