Overview
The real-time audiovisual translation model in the Qwen3.8 family, using a Hybrid-MoE Thinker–Talker design for speaker separation, synchronized bilingual output, and contextual disambiguation.
Qwen3.8-LiveTranslate Flash Realtime
The real-time audiovisual translation model in the Qwen3.8 family, using a Hybrid-MoE Thinker–Talker design for speaker separation, synchronized bilingual output, and contextual disambiguation.
QwenCloud provides `qwen3.8-livetranslate-flash-realtime` through the WebSocket Realtime API; it supports audio/image input, text/audio output, 60 source languages, and 29 audio-output languages.
MODEL NUMBER099
- First public date
- 2026-09-17
- Organization
- Alibaba Cloud
- Family
- Qwen
01
Specifications
Only published specifications are shown.
- Architecture
- Hybrid
- Input modalities
- Audio · Image
- Output modalities
- Text · Audio
- Open weights
- No
- Status
- Current
02
Overview
Why it mattered
It extends Qwen3.8 from general omni-modal and agent workflows into a dedicated low-latency speech-translation branch.
03
Architecture & capabilities
Attention
Hybrid-MoEThinker–TalkerInterleave
Model capabilities
Real-time audioNative audioSpeaker diarizationMultilingualLong context
04
Release history
QwenCloud released qwen3.8-livetranslate-flash-realtime with audio/image input, text/audio output, 60 source languages, 29 audio-output languages, and WebSocket Realtime API access.
View source05
Related models
Select a card to view the related model.
06
Sources
Suggest a correction
Copy this template, add the field, proposed value, and primary source, and share it with the archive maintainer.