Model release · Tencent
Hy-SimulMT
- Released
- Review updated
Tencent Hunyuan open-sourced Hy-SimulMT, a real-time Chinese–English simultaneous interpreting model that translates as the speaker talks, outputs text and speech, and keeps the speaker's voice via real-time voice cloning. Tencent cites low latency (~3s) and says it “Outperforms most open-source models and commercial APIs.”
Availability
Open source on Hugging Face, ModelScope and GitHub, with the SimulInfer inference service and SimulTrain training tools. Speech output and voice cloning need a configured TTS service; also available in the Hy Translate App. Repo license is listed as Other.