Instructions to use Qwen/Qwen3.8-27B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use Qwen/Qwen3.8-27B with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("image-text-to-text", model="Qwen/Qwen3.8-27B") 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("Qwen/Qwen3.8-27B") model = AutoModelForMultimodalLM.from_pretrained("Qwen/Qwen3.8-27B", 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]:])) - Inference
- HuggingChat
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use Qwen/Qwen3.8-27B with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "Qwen/Qwen3.8-27B" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Qwen/Qwen3.8-27B", "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/Qwen/Qwen3.8-27B
- SGLang
How to use Qwen/Qwen3.8-27B 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 "Qwen/Qwen3.8-27B" \ --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": "Qwen/Qwen3.8-27B", "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 "Qwen/Qwen3.8-27B" \ --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": "Qwen/Qwen3.8-27B", "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 Qwen/Qwen3.8-27B with Docker Model Runner:
docker model run hf.co/Qwen/Qwen3.8-27B
Deploy Qwen 3.8 27B private ai server easily
1
#253 opened 4 days ago
by
mzubairtahir000714
Update README.md
#252 opened 5 days ago
by
HOUNGYI
Add ParseBench-2 evaluation results
❤️ 1
2
#245 opened 8 days ago
by
boyang-runllama
MakoBench results: Total 13.8 (suite 4.1)
#244 opened 9 days ago
by
MakoBench
Infinite tool call loops for using Qwen 3.8 27B thinking mode off
1
#243 opened 10 days ago
by
imkenLuo
Getting rid of frp screen on phone
1
#242 opened 11 days ago
by
angelkuldeep
Keep the pre-filled thinking block out of the generation tags
👍 1
#241 opened 12 days ago
by
Rocketknight1
chat_template: OpenAI-style string tool_call.arguments still crashes with TypeError on |items
👍 1
#240 opened 12 days ago
by
SamJoshua
Add MultiLoKo results in .eval_results/
#239 opened 15 days ago
by
dieuwkehupkes
Please, make also 27B A3B !
👍 16
2
#238 opened 17 days ago
by
ZeroWw
Qwen3.8-9B someday?
👀👍 9
2
#237 opened 17 days ago
by
nerun
Add IFBench evaluation result
2
#235 opened 18 days ago
by
SaylorTwift
Notes from an independent NVFP4 port of Qwen3.8-27B
👍 2
1
#234 opened 18 days ago
by
alexmv2025
Create Manmen
#233 opened 19 days ago
by
Manmen58
Upload 20240618_213433.jpg
😔 2
#232 opened 20 days ago
by
Jeevan9855
the j-lens on this model is better than the reasoning trace
❤️👍 2
#231 opened 21 days ago
by
john1248
Add Terminal-Bench evaluation results
👍❤️ 2
#230 opened 22 days ago
by
nielsr
Qwen in production: auto-generating 79 baseball highlight clips for a Korean broadcaster at the 2026 Asian Games
#228 opened 23 days ago
by
abafffff
Local GPU benchmark on 2x RTX 5060 Ti (32GB, vLLM 0.29): 64.5% overall (4 benchmarks) at 60.7 tok/s
#225 opened 24 days ago
by
Mausbiber
Upload 1679313310.jpg
#224 opened 25 days ago
by
Vlad935
Why only focusing on 27 and 35 ????
👍 2
4
#223 opened 25 days ago
by
nenor
Add OmniCompass evaluation results
#222 opened 26 days ago
by
VCLab-PolyU
Feature Request: Qwen3.8-22B + 30B N-gram Embeddings — Optimized for 16GB GPUs ------- 标题:功能建议:推出 Qwen3.8-22B + 30B N-gram Embedding 版本,专为 16GB 显存 GPU 优化
🧠❤️ 2
1
#221 opened 27 days ago
by
HALO117MC
Quen model
#220 opened 27 days ago
by
Himabinduvelamakuri
Tool failures 1 of 10. Anyone having this issue? Fixes?
#219 opened 29 days ago
by
inferencia
Is there an official list of supported languages?
#218 opened about 1 month ago
by
NSMathpati
Add Modality Maturity Index results
#217 opened about 1 month ago
by
dieuwkehupkes
Is there any Official fix for the Chat-Template for Claude Code ?
#216 opened about 1 month ago
by
aygupt
Qwen3.8-27B-Flash-MTP-GGUF--5.79 GB. Code ~70-90 t/s, editing ~100-200 t/s.
#213 opened about 1 month ago
by
SuperLogic
Intermittent Persian language-instruction failures after tool results — Alibaba via OpenRouter
#212 opened about 1 month ago
by
Amin2747
谢谢Qwen 送了个万能老婆给我
👀 3
#211 opened about 1 month ago
by
iceair323
New MEDIUM MoE model
🚀 5
#210 opened about 1 month ago
by
bakrhere57
Please, please, release a 35b MOE version of 3.8
🚀 12
1
#209 opened about 1 month ago
by
jzed205
jailbreak mode?
#208 opened about 1 month ago
by
Umbron
Idea for a future Qwen 27B: nine 1M-token agents on one RTX 5090
1
#206 opened about 2 months ago
by
BigBlueWhale
Add official Agents Last Exam evaluation results
❤️ 1
#205 opened about 2 months ago
by
sunyiyou
testing
#204 opened about 2 months ago
by
mdshahedrahman
Update chat_template.jinja
1
#203 opened about 2 months ago
by
mdshahedrahman
Harness runs Qwen 3.8 27B for 24hr on an RTX5090, installs software and writes a Postgres-SpringBoot-React spreadsheet application fully unattended
❤️ 1
3
#202 opened 2 months ago
by
anglepoiselife
Qwen 3.8 27b vram requirements
1
#201 opened 2 months ago
by
mzubairtahir000714
Beta testers wanted (Qwen 3.8:27b)
8
#199 opened 2 months ago
by
anoneemoose
really great job
#197 opened 2 months ago
by
llMss12
Open-weight policy is moving against this model's weights: the 'apply the same standards' argument, in their own words
❤️👍 3
3
#196 opened 2 months ago
by
SmilingQbandit
New Top #1 most liked on HF! 🤗
🤗 9
#195 opened 2 months ago
by
FalconNet