Skip to content
AI HOT (Curated Pool)

Qwen launches Qwen3.8-LiveTranslate real-time interpretation model with 2.3s latency and speaker separation

Qwen 发布 Qwen3.8-LiveTranslate 实时同传模型,LAAL 降至 2.3 秒

Qwen3.8-LiveTranslate cuts simultaneous interpretation latency from 2.8s to 2.3s by interleaving audio and text into a single stream. It supports 60 input languages, real-time speaker separation with voice cloning, and synchronized bilingual output. On the Omnilingua-MSpeaker benchmark it outperforms current mainstream systems in faithfulness, fluency, and conciseness. API is available via Alibaba Cloud DashScope.

Why it matters: Qwen3.8-LiveTranslate cuts average simultaneous interpretation latency from 2.8 seconds to 2.3 and adds speaker separation, so each turn keeps its own voice.

Read the original ↗Export Markdown