Qwen3.8-27B
A 27B dense multimodal model under Apache 2.0 that reads images and video, reasons step by step and fits on one GPU.
In short. Qwen3.8-27B is a mid-sized AI model from Alibaba that you can run on one good graphics card, or on a high-memory laptop once it is compressed. It understands text, pictures and video, can think through problems step by step, and its license lets you use it commercially without asking.
Summary
Qwen3.8-27B is the compact member of Alibaba's Qwen3.8 generation. Alibaba published the weights on 14 August 2026 under the Apache 2.0 license. It is a dense model with 27.8 billion parameters, so every parameter is used for every token, which keeps memory needs predictable.
It is a native vision-language model: it accepts text, images and video and writes text. The context window is 262,144 tokens natively and can be stretched to 1 million. Thinking mode is on by default and can be switched off per request, and a reasoning-effort setting controls how long the model thinks. The layers mix Gated DeltaNet, a linear-attention design, with ordinary gated attention in a 3-to-1 pattern, which keeps long inputs cheaper than in a plain transformer.
Alibaba says the model matches Qwen3.7-Plus, a mixture-of-experts model about ten times its size, on its own evaluations; this is a vendor claim. The BF16 checkpoint is about 56 GB and the official FP8 checkpoint about 31 GB. A 4-bit community quantization needs roughly 17 to 20 GB for the weights, which is within reach of a 24 GB graphics card.
Good for
- A strong all-round assistant on a single 24 to 48 GB GPU or a Mac with enough memory
- Coding and agent workflows where you want a commercially usable license
- Reading screenshots, documents and video clips locally
- Fine-tuning, since dense models are simpler to train than very large mixture-of-experts models
Watch out for
- Quality is below the 2.4T Qwen3.8-Max, and the matching claim against Qwen3.7-Plus is Alibaba's own
- Long contexts need much more memory than the weights alone; plan for extra headroom
- It writes text only and cannot generate images, speech or video
- Check the exact quantized build you download, because quality varies between community versions
Performance claims come from the developer and have not been independently verified.
How much memory does it need?
Estimates for the 27B variant. A mixture-of-experts model must hold all of its experts in memory even though only some are used for each token, so total parameters decide the memory and active parameters decide the speed.
| Variants | Parameters | BF16 | FP8 / INT8 | 4-bit (Q4) | Fits on |
|---|---|---|---|---|---|
| 27B BF16 | 27.8B | 64 GB | 32 GB | 19 GB | Gaming GPU |
| 27B FP8 FP8 | 27.8B | 64 GB | 32 GB | 19 GB | Gaming GPU |
Estimates: weights plus about 15% overhead. Real usage grows with context length and depends on the software.
The license in plain words
One of the most widely used permissive licenses. You may use, change and sell products built on the model with only light conditions.
See also
- Gemma 4 vs Qwen3.8-27B: which open model should you run on one GPU?
- Best open-source LLMs in 2026
- Best open-source models for coding and AI agents
- Best open-source multimodal models in 2026 (images, video and audio)
- Apertus 1.5: Switzerland's fully open model reads images and audio
- Clef: Cloudflare's open decision models for agent workflows
- Nanbeige4.3-33B: a 33B agent model with a 23 GB 4-bit build
- PrismML's Ternary Bonsai 2: a 27B-class model in 5.95 GB
- Qwen3.8: three language models, three licenses, and a fourth for images
Frequently asked questions
- What is Qwen3.8-27B?
- Qwen3.8-27B is a mid-sized AI model from Alibaba that you can run on one good graphics card, or on a high-memory laptop once it is compressed. It understands text, pictures and video, can think through problems step by step, and its license lets you use it commercially without asking.
- Can I use Qwen3.8-27B commercially?
- Yes. Qwen3.8-27B is released under Apache 2.0, a permissive license. You can use it commercially if you follow the standard conditions, such as keeping the license notice.
- Is Qwen3.8-27B open source?
- Qwen3.8-27B is released under Apache 2.0, a license approved by the Open Source Initiative. Training data and training code may not be public, so it is best described as open weights under an open-source license.
- Can I run Qwen3.8-27B on my own computer?
- Yes, on a good graphics card. The smallest variant, 27B, needs about 19 GB of memory at 4-bit precision, which suits a 24 to 32 GB gaming GPU or a Mac with 32 GB or more.
- When was Qwen3.8-27B released?
- Qwen3.8-27B was released by Alibaba Qwen on August 14, 2026.
- How long a context window does Qwen3.8-27B have?
- Qwen3.8-27B supports a context window of up to 256K tokens.
Official links
Sources
Specifications come from these primary sources. Benchmark claims, where mentioned, are reported by the developer.
- Qwen3.8-27B model card (Hugging Face) huggingface.co
- Alibaba Unveils Qwen3.8-27B and Releases Weights of Qwen3.8 Flagship Model (Alibaba Cloud, 17 Aug 2026) alibabacloud.com
- Qwen3.8 repository and release news (GitHub) github.com
Release date and facts last verified November 8, 2026.
Related models
| Model | Released | Parameters | Context | License | Smallest variant needs |
|---|---|---|---|---|---|
| Gemma 4 Google DeepMind | 5.1B – 30.7B | 256K | Apache 2.0 (Permissive) | 3.52 GB Laptop or phone | |
| Qwen3.8-Flash-Next Alibaba Qwen | 180B 6B active | 256K | Qwen Community 1.0 (Conditional) | 124 GB Big-memory desktop | |
| Olmo 3.1 Allen Institute for AI (Ai2) | 7.3B – 32.2B | 64K | Apache 2.0 (Permissive) | 5.04 GB Laptop or phone | |
| Nanbeige4.3-33B Nanbeige LLM Lab | 33.1B | 512K | Apache 2.0 (Permissive) | 23 GB Gaming GPU |