Tongyi Lab (Alibaba) · Voice Agents
Wan Streamer v0.2
End-to-end Transformer for real-time duplex audio-visual conversation that perceives and responds with synchronized speech and video as a single causal stream; v0.2 raises output resolution from 192x336 to 640x368 at 25 FPS while holding ~200ms model-side (~550ms total) latency via a Thinker-Performer architecture, and can embody any natural-language-described character.
Open the source ↗Similar Audio & Speech releases
- Dessa
Block (Dessa research lab) · Speech-only conversation model · Week 45, 2026
- Eleven v4 and v4 Turbo
ElevenLabs · Expressive TTS Model · Week 41, 2026
- Gemini 3.8 Live
Google · Live Voice Agent · Week 39, 2026
- VoiceMem
Tsinghua University · Voice Memory · Week 35, 2026
- Gemini 3.5 Transcribe
Google · Speech Recognition · Week 35, 2026
- Gemini Mac Voice
Google · Voice Agents · Week 31, 2026
More from Tongyi Lab (Alibaba)
- Wan-Dancer
Tongyi Lab (Alibaba) · Text-to-Video · Week 29, 2026
- VACE
Tongyi Lab (Alibaba) · Video Editing · Week 19, 2025
Published by Quanta AI Labs, an AI agency in India. Browse every week in AI News Weekly.