The release and what it targets
Alibaba released Qwen3.8-27B, a 27-billion-parameter open-weight artificial intelligence model designed to run on consumer hardware such as laptops, entering an on-device AI market that Meta entered the prior week with its own laptop-ready model. Alibaba said the model is suited to coding, professional work, research and what the company described as long-horizon agentic tasks, and that it matches the performance of a model ten times its size.
Direct competition with Meta's Muse Glimmer
The launch comes days after Meta introduced Muse Glimmer, a 30-billion-parameter open-weight model built to run on a Mac or PC with a single consumer graphics card, and made it available for free download under an Apache 2.0 license. Meta also said it intends to make the weights of Muse Spark 1.2, its flagship model, publicly available. Qwen3.8-27B sits in roughly the same parameter range and is pitched at the same single-machine use case, positioning Alibaba as a direct entrant into a category Meta had just opened.
A broader open-weight push
Alongside the laptop-focused model, Alibaba released the weights of Qwen3.8-Max, its most capable model, making it available for free download. Alibaba said Qwen3.8-Max scales to 2.4 trillion total parameters with 95 billion active parameters, calling it the first time the company has open-sourced a model at that scale. Separately, the company is reported to be planning to require large commercial users of the open-weight version of Qwen3.8-Max to share a portion of their revenue.
Where Alibaba stands in the open-weight race
Alibaba has built a leading position in the open-weight AI field, with Hugging Face reporting last week that models built on Qwen have generated 151,448 derivative works, putting Alibaba's tally at roughly 2.6 times that of Meta. Nick Patience, AI lead at the Futurum Group, said Meta's re-embrace of open weights was itself a response to two years of Chinese labs taking a large share of the open-weight market, underscoring the competitive dynamic the new releases are designed to address.
What remains uncertain
The reporting does not specify which benchmark Alibaba used to claim Qwen3.8-27B matches a model ten times its size, and details on the planned revenue-sharing terms for large commercial users of Qwen3.8-Max are not described in the available sources. Adoption metrics for the newly released Qwen3.8-27B weights are not yet available.
Share this article
