Tencent Open Sources WeMM-Embedding: 9B Model Ranks #1 on MMEB v2/v3

aigclink · x · 2026-08-26

Tencent's WeChat Vision Team has open-sourced the WeMM-Embedding family of multimodal embedding models. Supporting unified representations for text, images, videos, visual documents, and interleaved inputs, it powers search and recommendations across WeChat Channels, Official Accounts, Moments, and e-commerce. Available in 2B, 4B, and 9B sizes, the 9B model ranked #1 on both MMEB-v2 and MMEB-v3 benchmarks.

Related event: Tencent Open-Sources WeMM-Embedding Multimodal Model Family, Topping MMEB Leaderboards(7 posts)→

Original post →

More from Multimodal

Multimodal channel →