Tencent open-sources WeMM multimodal embedding models, 9B ranks first on dual benchmarks
aigclink · x · 2026-08-26
Tencent's WeChat Vision team open-sourced the WeMM-Embedding family of universal multimodal embedding models, available in 2B, 4B, and 9B sizes. The models support text, image, video, and interleaved inputs for unified retrieval. The 9B version ranked first on both MMEB-v2 and MMEB-v3 benchmarks and is now available on GitHub and Hugging Face.
Related event: Tencent Open-Sources WeMM-Embedding Multimodal Models, 9B Tops MMEB v2/v3(7 posts)→
More from Models
- Grok Voice 3.0 Tops Design Arena Speech Leaderboard, Up 7.4 Points — XFreeze · 2026-10-16
- Robotics LLMs tested on 16 tasks: GPT-6 Astra hits 17.2% success vs Fable's 9.1% — pabbeel · 2026-10-16
- Meta open-sources Muse Spark: top US open model, scores just below DeepSeek — bindureddy · 2026-10-16
- TRL adds day-0 support for fine-tuning Muse Spark Open's 816B MoE with LoRA on 64 H100s — QGallouedec · 2026-10-16
- Meta's Muse Spark Open ships detailed eval docs; researcher pushes machine-readable EEE standard — evijit · 2026-10-16
- Baseten serves Meta's Muse Spark Open on day 0 at $1.25/M input tokens — baseten · 2026-10-16