Alibaba releases Qwen3.8-27B, open-weight AI model for laptops

Alibaba released Qwen3.8-27B on Monday, an open-weight AI model engineered to run on consumer laptops and desktop computers, marking the latest salvo in the intensifying competition for on-device artificial intelligence.

The model contains 27 billion parameters and is designed for coding, professional work, research, and what Alibaba describes as long-horizon agentic tasks. According to the company, Qwen3.8-27B delivers performance comparable to models ten times its size.

The launch arrives less than a week after Meta introduced Muse Glimmer, a 30-billion-parameter open-weight model built to run on Mac or PC hardware with a single consumer graphics card. Meta released Muse Glimmer under an Apache 2.0 license for free download.

Alibaba opens its flagship model

Alongside the laptop-focused release, Alibaba made the weights of Qwen3.8-Max available for free download. The company said Qwen3.8-Max scales to 2.4 trillion total parameters with 95 billion active parameters — the first time Alibaba has open-sourced a model at that scale.

The move extends Alibaba's aggressive push into open-weight AI development. Hugging Face reported last week that models built on Qwen have generated 151,448 derivative works — third-party products built on top of the downloaded weights — roughly 2.6 times Meta's tally.

Nick Patience, AI lead at the Futurum Group, noted that Meta's renewed commitment to open weights represents a response to Chinese labs capturing significant market share in the open-weight space over the past two years.

Why it matters

The on-device AI market represents a strategic battleground for model developers. Locally run models can deliver faster responses and keep sensitive data off remote servers, addressing privacy concerns that matter to enterprise and individual users. As major tech companies race to make powerful AI accessible on consumer hardware, the open-weight approach allows broader experimentation and commercial deployment without the infrastructure costs of cloud-based inference. Alibaba's dominance in derivative works suggests its models have become a foundation for third-party AI applications, potentially shaping the broader ecosystem.

Commercial licensing in development

Alibaba released Qwen3.8-Max earlier in August through its QwenCloud platform. Following that announcement, the company's stock rose 4.5% in premarket trading in New York and 7% on the Hong Kong exchange. Separately, Alibaba is reportedly planning to require large commercial users of the open-weight version of Qwen3.8-Max to share a portion of their revenue.

The details were first reported by Quartz.