Tencent's WeMM-Video pushes open-source omni-modal video understanding
tencent · hf · 2026-11-20
Tencent WeChat Vision and Tsinghua released WeMM-Video, a family of Chinese-English bilingual omni-modal foundation models unifying text, image, audio, and video for joint audio-visual perception and reasoning.
- Sizes span 2B, 4B, 9B, and 35B-A3B, with a hybrid thinking paradigm: fast non-thinking responses plus step-by-step thinking mode
- Trained on roughly 1.2 trillion multimodal tokens with progressive pre-training and multi-stage post-training, achieving strong open-source video understanding results
More from Multimodal
- Genie 4 users build playable worlds in 30 seconds; Jurassic Park world ejects you from the car — iamfakhrealam · 2026-11-20
- Genie 4 demo: 20-second Miami police chase generated live at 24fps with stereo audio — iamfakhrealam · 2026-11-20
- Google opens Genie 4 to regular users: single-prompt playable worlds go viral — iamfakhrealam · 2026-11-20
- Alibaba's TaoTex diffusion model generates native 3D materials, beating TRELLIS.2 — jiqizhixin · 2026-11-20
- AI turns Mediterranean illustration into interactive 3D world; Three.js wall-crash demo costs $6.98 — Saboo_Shubham_ · 2026-11-20
- Kling team's Coda-DaVinci reimagines video generation via Code2Video agents — KlingTeam · 2026-11-20