Tencent, SenseTime announce new AI models to rival Google, Anthropic
What's the story
Tencent and SenseTime have unveiled their latest artificial intelligence (AI) models at the World Artificial Intelligence Conference (WAIC) in Shanghai. The move marks a major development in the ongoing rivalry among Chinese tech giants. Tencent, a leading player in social media and gaming, introduced its Hunyuan 3D World Model 1.0. The open-source AI model can create complex three-dimensional environments from natural language or image inputs.
Innovation
Hunyuan generates interactive virtual 3D scenes
Tencent's new Hunyuan model can generate interactive, 360-degree virtual 3D scenes. This capability could greatly simplify the process of creating virtual reality experiences and video games. The tech giant also touted its latest model as the industry's first open-source, 3D world-generation AI that is fully compatible with "CG pipelines." They are standard workflows for creating 3D graphics and animations in film production, gaming, and visual effects.
Competition
SenseNova V6.5 outperforms Gemini 2.5 Pro
AI pioneer SenseTime also unveiled its latest model, SenseNova V6.5, at the WAIC event. The Hong Kong-listed firm's Chairman and CEO Xu Li claimed that their new model has outperformed some of its US counterparts such as Google's Gemini 2.5 Pro and Anthropic's Claude 4-Sonnet. The launch comes months after the release of SenseNova V6 in April, a multimodal model that had already beaten OpenAI's GPT-4o on several fronts.