Alibaba announced its intention to unveil an open-source version of its video and image-generating artificial intelligence model, Wan 2.1, in a post on X on Tuesday. The company will provide comprehensive details in a recorded video at 11 p.m. (1500 GMT), according to a spokesperson.
The release of this model comes amid increasing competition in China's AI market, following the launch of its previous open-source models last month. These models have made a significant impact on the global AI industry by matching the performance of top products from companies like OpenAI, which has started focusing on closed-source offerings.
Initially introduced in January, Alibaba's latest video and image-generating AI model was later rebranded from Wanx to Wan, emphasizing its capacity to produce highly realistic visuals. Alibaba previously emphasized its leading position on VBench, a video generative models leaderboard, particularly in critical aspects such as multi-object interactions.
Recently, Alibaba introduced a preview version of its reasoning model, QwQ-Max, with plans to make it open source upon the full version's official release, scheduled soon.
Alibaba also disclosed its commitment to investing a minimum of 380 billion yuan ($52 billion) over the next three years to strengthen its cloud computing and AI infrastructure earlier this week.