AIQwen3.8-LiveTranslate debuts with synchronized bilingual display
Qwen says its real-time interpretation model reduces average lagging from 2.8 to 2.3 seconds across 60 languages while improving faithfulness, fluency and conciseness.
TLDR
Qwen announced Qwen3.8-LiveTranslate, a real-time simultaneous interpretation model built on an Interleave architecture. Qwen says it can distinguish speakers in multi-party conversations and preserve each speaker’s voice through more stable voice cloning. Other announced features show the source and translation together on screen and use conversation history to clarify names and terminology for consistent translations.
Combined views
229.7K
1 Source, first seen 1d ago
Qwen3.8-LiveTranslate debuts with synchronized bilingual display
Qwen says its real-time interpretation model reduces average lagging from 2.8 to 2.3 seconds across 60 languages while improving faithfulness, fluency and conciseness.
TLDR
Qwen announced Qwen3.8-LiveTranslate, a real-time simultaneous interpretation model built on an Interleave architecture. Qwen says it can distinguish speakers in multi-party conversations and preserve each speaker’s voice through more stable voice cloning. Other announced features show the source and translation together on screen and use conversation history to clarify names and terminology for consistent translations.