Skip to main content
Watch: I Built an AI App Without Leaving Databricks

Alibaba Cloud (Qwen) - Dense - verified

Run Qwen3.5-4B locally

Qwen3.5-4B is a 4.7B dense model from Alibaba Cloud (Qwen) with a 262K-token context window, released under Apache License 2.0 (commercial use is allowed); the smallest listed download is 2.7 GB, the Q4_K_M GGUF build from unsloth/Qwen3.5-4B-GGUF.

Official weights: Qwen/Qwen3.5-4B - modalities: text, image, video

Total params
4.7B
Architecture
Dense
Context
262K
Smallest download
2.7 GB

Fits in

Smallest download at most 85% of memory

Downloads

The official weights and every quantization listed for this model, with file sizes from Hugging Face.

Qwen3.5-4B downloads and file sizes
DownloadFormatSizeRepo
Official weightsBF169.3 GBQwen/Qwen3.5-4B
Q4_K_MSmallest listedGGUF2.7 GBunsloth/Qwen3.5-4B-GGUF
Q8_0GGUF4.5 GBunsloth/Qwen3.5-4B-GGUF
4bitMLX3.0 GBmlx-community/Qwen3.5-4B-4bit

GB is decimal: 1 GB = 10^9 bytes. Add memory for the context and the operating system on top of the file size.

Run it

Copied verbatim from the page linked under each command. Server commands from model cards often assume a multi-GPU node; check the flags before you run them.

Ollama

Terminal
ollama run qwen3.5:4b

Source: ollama.com/library/qwen3.5:4b

vLLM

Terminal
vllm serve Qwen/Qwen3.5-4B --port 8000 --tensor-parallel-size 1 --max-model-len 262144 --reasoning-parser qwen3

Source: huggingface.co/Qwen/Qwen3.5-4B

SGLang

Terminal
python -m sglang.launch_server --model-path Qwen/Qwen3.5-4B --port 8000 --tp-size 1 --mem-fraction-static 0.8 --context-length 262144 --reasoning-parser qwen3

Source: huggingface.co/Qwen/Qwen3.5-4B

Transformers

Terminal
transformers serve --force-model Qwen/Qwen3.5-4B --port 8000 --continuous-batching

Source: huggingface.co/Qwen/Qwen3.5-4B

Not sure which runtime to use? Compare local LLM runtimes by platform, API and features.

License

License
Apache License 2.0apache-2.0
Commercial use
Commercial use: YesCommercial use is allowed.
Gated download
No. The files download directly.

Worth knowing

Thinking mode is on by default; context is 262,144 tokens natively and extensible to 1,010,000 with YaRN.

Released (Hugging Face repo creation date).

FAQ

How big is the Qwen3.5-4B download?

The official BF16 weights in Qwen/Qwen3.5-4B are 9.3 GB. The smallest listed download is 2.7 GB, the Q4_K_M GGUF build from unsloth/Qwen3.5-4B-GGUF. Plan for more memory than the file size: the context (KV cache) and the operating system need room too.

Can I use Qwen3.5-4B commercially?

Qwen3.5-4B is released under Apache License 2.0, and commercial use is allowed. The license text is at https://huggingface.co/Qwen/Qwen3.5-4B/blob/main/LICENSE.

How much memory do I need for Qwen3.5-4B?

By this directory's rule of thumb it fits in 8 GB and up: the smallest listed download, 2.7 GB, is at most 85% of 8 GB, which leaves room for the context and the operating system.

How do I run Qwen3.5-4B with Ollama?

Run "ollama run qwen3.5:4b", as shown at https://ollama.com/library/qwen3.5:4b. The Ollama library lists the qwen3.5:4b tag at 3.3GB - 4.0GB.

What is the context length of Qwen3.5-4B?

262,144 tokens. A longer context needs more memory for the KV cache, so you can run with a shorter one when memory is tight.

Sources

Last verified from the pages below.

Newsletter

New open weights, every week

One email per week. New model releases, open-weight drops and what they mean for what you build.

Keep exploring

Best local AI models

Our curated picks by hardware tier, with benchmarks and the reasoning.

Model API pricing and specs

First-party prices per 1M tokens when you would rather call an API.

AI benchmark leaderboards

Sourced scores, for when you need to compare quality.

Setup guides

Point your coding tools at the model you picked.