Tencent Hunyuan merges multimodal and LLM teams: A key pivot in technical roadmap

In Q2 2026, Tencent's Hunyuan large model call volume grew by over 180% quarter-over-quarter. Multimodal models are seeing significantly faster commercialization in niche areas like image generation and 3D modeling compared to pure text dialogue. However, right at this seemingly triumphant moment, Tencent internally completed a low-key but far-reaching organizational adjustment affecting R&D direction—on July 23, Tencent announced the merger of the Hunyuan multimodal model department with the large language model department to form a Foundation Model Department, unified under the management of Tencent's Chief AI Scientist, Yao Shunyu. The signals released by this move are far more worth digging into than the numbers in a quarterly earnings report.