tencent
3 posts on tencent, newest first.
- Speech models, image tools and video editingNew open speech-to-text and voice models, image models for layouts and vector art, and tools that edit video for you.Issue 6
- Genie 4, robot training worlds and image-to-3DGoogle DeepMind's Genie 4 adds sound to generated worlds, new tools train robots in simulation, and several models turn one image into a 3D object.Issue 4
- Small open models for coding, vision and OCRTen new open models, from a 1 GB text model to a 35B research model, plus new tools for vision and long documents.Issue 2