Tencent released Hy3, an open-source language model built on a mixture-of-experts architecture with 295 billion total parameters and 21 billion active at any time.
The company says Hy3 can match models two to five times its active size and cut its hallucination rate to 5.4 percent. Those claims will need independent testing, but the architecture choice reflects the industry’s push for efficiency.
The release adds another major Chinese model to the open ecosystem at a time when developers are comparing not only benchmarks, but also deployment cost and inference efficiency.