Alibaba Qwen Team Introduces Qwen3.5-LiveTranslate-Flash: Real-Time Multimodal Interpretation Across 60 Languages at 2.8-Second Latency
2.8 seconds. That's how fast Alibaba's new Qwen3.5 translates across 60 languages in real-time—outperforming commercial alternatives on FLEURS and CoVoST2.
Why it matters
Alibaba's multimodal translation model demonstrates a significant capability leap in real-time speech-to-speech with vision enhancement, challenging OpenAI/Google's translation dominance and enabling enterprise deployment across 60 languages.
The key facts
7 to knowQwen3.5-LiveTranslate-Flash processes audio and video simultaneously
60 input languages, 29 output speech languages
2.8-second latency for end-to-end translation
Outperforms commercial alternatives on FLEURS and CoVoST2 benchmarks
Features real-time speaker voice cloning and lip-reading vision enhancement
Available via Alibaba Cloud Model Studio API with WebSocket protocol
Domain-specific keyword configuration support
Go to the source
MarkTechPostmarktechpost.com
Publisher excerpt: Alibaba's Qwen team has released Qwen3.5-LiveTranslate-Flash, a real-time multimodal translation model that processes audio and video simultaneously. The model covers 60 input languages and produces speech output in 29 languages at 2.8 seconds of latency. Key additions over the previous Qwen3…