Tencent launches an intelligent agent development platform and plans to open source multiple models

According to Zhitong Finance APP, on May 21, Tencent Cloud announced at the "2025 AI Industry Application Summit" that its large model knowledge engine has been upgraded to the Tencent Cloud Intelligent Agent Development Platform (TCADP) and officially released, integrating DeepSeek-R1, V3 models, and online search. The platform aims to assist enterprises and developers in building intelligent solutions more efficiently, promoting the widespread application of AI technology across various industries. In addition, Tencent plans to launch a world model mixed-element 3D scene model soon and open-source enterprise-level mixed reasoning models, edge-side mixed reasoning models, and multimodal foundational models.

Recently, Tencent's large model products have undergone a batch update, introducing the mixed-element T1 Vision model for visual deep reasoning, the mixed-element Voice model for end-to-end voice calls, and the mixed-element Image 2.0 model, among others. The Tencent mixed-element large model series continues to iterate and upgrade, with models like the mixed-element T1 Preview model and the Turbo S model already applied in several internal products. Tencent's mixed-element large models actively embrace open-source, covering multiple modalities such as text, images, video, and 3D generation. On May 16, Tencent released the mixed-element Image 2.0 model, which, based on innovative model architecture, achieves real-time live image generation, providing a new AI live image interaction experience. Previously, Tencent also open-sourced the mixed-element text-to-image model and the mixed-element 3D model, among others