Tencent open-sources Hy-SimulMT: 30B/A3B realtime translation with 3s lag
AdinaYakup · x · 2026-10-26
Tencent released Hy-SimulMT, a realtime simultaneous translation model: 30B total / A3B active params, Apache 2.0, Chinese<>English, 3-second lag (lowest in its RealSI eval), with GGUF and fp8 builds for local deployment. The benchmark chart places it top-left (high GEMBA quality, low StreamLAAL latency), ahead of Qwen3.8-LiveTranslate, GPT-Realtime-Translate, Seed-LiveInterpret 2.0, and Gemini3.5-LiveTranslate.
More from Models
- Fable 5.5 launch rumors persist for third week, but no frontier model drops this month — patience_cave · 2026-10-26
- Abliterated Model Flops in the Wild: It Got Worried About My Mental Health Instead — ryunuck · 2026-10-26
- OpenAI upgrades Decisions API: 400 questions per call, native voice and image support — OpenAIDevs · 2026-10-26
- Fangcun AI, China's first independent frontier AI safety lab, launches with DeepSeek eval report — JacquesThibs · 2026-10-26
- When AI hits an existential crisis: model confuses itself and everyone else — Ok-Delivery-7437 · 2026-10-26
- Kimi K3.1 Pro tops MathArena open models, closing frontier gap to about 6 months — teortaxesTex · 2026-10-26