Instructions to use TK-13/Swift1.5-Qwen3.8-Flash-Next-FP8 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use TK-13/Swift1.5-Qwen3.8-Flash-Next-FP8 with Transformers:

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

pipe = pipeline("image-text-to-text", model="TK-13/Swift1.5-Qwen3.8-Flash-Next-FP8")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("TK-13/Swift1.5-Qwen3.8-Flash-Next-FP8")
model = AutoModelForMultimodalLM.from_pretrained("TK-13/Swift1.5-Qwen3.8-Flash-Next-FP8", device_map="auto")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.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=256)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use TK-13/Swift1.5-Qwen3.8-Flash-Next-FP8 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "TK-13/Swift1.5-Qwen3.8-Flash-Next-FP8"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "TK-13/Swift1.5-Qwen3.8-Flash-Next-FP8",
		"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/TK-13/Swift1.5-Qwen3.8-Flash-Next-FP8

SGLang

How to use TK-13/Swift1.5-Qwen3.8-Flash-Next-FP8 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 "TK-13/Swift1.5-Qwen3.8-Flash-Next-FP8" \
    --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": "TK-13/Swift1.5-Qwen3.8-Flash-Next-FP8",
		"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 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 "TK-13/Swift1.5-Qwen3.8-Flash-Next-FP8" \
        --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": "TK-13/Swift1.5-Qwen3.8-Flash-Next-FP8",
		"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"
						}
					}
				]
			}
		]
	}'

Docker Model Runner
How to use TK-13/Swift1.5-Qwen3.8-Flash-Next-FP8 with Docker Model Runner:
```
docker model run hf.co/TK-13/Swift1.5-Qwen3.8-Flash-Next-FP8
```

Swift1.5-Qwen3.8-Flash-Next-FP8

An unofficial FP8 quantization of ukisai/Swift1.5-Qwen3.8-Flash-Next in the exact layout and recipe of Qwen/Qwen3.8-Flash-Next-FP8 (128 x 128 block FP8 routed experts, BF16 scales, dynamic activations, no calibration). The recipe was verified by recreating Qwen's FP8 checkpoint bitwise from its BF16 release: TK-13/Qwen3.8-Flash-Next-FP8-Recreated. Not affiliated with or endorsed by UkisAI or Qwen.

Full model card (recipe, validation, evaluation) coming soon.

License

Swift Open License v1.0 (LICENSE, Copyright 2026 UkisAI; commercial use limited, see Section 5), on top of Qwen3.8-Flash-Next (Copyright (c) 2026 Qwen, Qwen Community License 1.0, LICENSE-QWEN). See NOTICE.

Changes made in this repo (TK-13, 2026-10-24): model-*.safetensors and model.safetensors.index.json (routed-expert and PLE n-gram-embedding weights quantized to FP8, fused experts split per expert); config.json (quantization_config added); README.md replaced; NOTICE appended. All other files are unmodified from ukisai/Swift1.5-Qwen3.8-Flash-Next.

Downloads last month: 93

Safetensors

Model size

180B params

Tensor type

BF16

F8_E4M3

I64

Model tree for TK-13/Swift1.5-Qwen3.8-Flash-Next-FP8

Base model

Qwen/Qwen3.8-Flash-Next

Finetuned

ukisai/Swift1.5-Qwen3.8-Flash-Next

Quantized

(56)

this model