Unverified50% confidenceFactExact time
Both models support controlling conversation through non-verbal cues such as <laughs>, <sigh>, and <gasp>, as well as active-listening back-channel markers such as |mhm| and |yeah|.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/23/2026
First Seen
Valid until: 12/22/2026
Sources
Related Entities
Related Claims
UnverifiedDia能根据输入文本自动生成两人对话音频，可模拟咳嗽声、笑声等非语言细节71% similarVerifiedEnd-to-end voice models take audio directly as input and output, performing understanding and generation at the audio token level while preserving acoustic features such as intonation, speaking rate, sighs, and laughter.71% similarUnverifiedThe new model can perceive paralinguistic information such as the speaker's volume and tone, and proactively raise its volume or switch states.70% similarUnverifiedThe workflow offers two voice generation modes: a lip-sync mode (which freezes the user's voice into the audio latent space) and a model-generated speech mode.68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/944525API
curl https://kongchang.com/api/v1/knowledge/claims/944525MCP
get_claim(id=944525)