Alibaba Group has released the new generation of its large language model that can understand text, audio, images and video. But this time, the Chinese tech giant is releasing the model, Qwen3.5-Omni, as a proprietary model—unlike its open-source predecessor. The fact that Alibaba decided not to open-source Qwen3.5-Omni signals a potential change in the company’s AI model strategy, which has focused heavily on open-source models in the past few years. Now, instead of allowing users to download the latest model for free, the company is focusing on generating more revenue from enterprise customers who will pay for it. Alibaba’s previous-generation Omni model released in September, Qwen3-Omni, was an open-source model.

Full briefing : Alibaba’s new Qwen3.5-Omni multimodal model, which processes text, audio, images, and video, is proprietary, marking a shift away from its open-source strategy.