Unverified50% confidenceFactExact time
Audio tokens are trained with the objective of reconstructing high-fidelity audio, retaining all acoustic information including timbre, emotion, and ambient sound, rather than preserving only linguistic content as ASR output does.
1
Sources
50%
Confidence
Long-term
Relevance
8/9/2026
First Seen
Sources
深度实测：ChatGPT语音模式能听懂语气、口音和情绪
Related Claims
UnverifiedEnd-to-end native architectures directly process audio input and output audio without relying on intermediate conversion steps, preserving richer acoustic features.76% similarUnverifiedThe new end-to-end voice model is trained directly at the audio level, without first converting speech to text, and can directly capture acoustic features and emotions.76% similarUnverified声音克隆通过说话人编码器将声音映射为高维嵌入向量，说话人编码器训练通常采用对比学习或三元组损失76% similarUnverified说话人嵌入向量捕捉声音中与内容无关的身份信息，包括声带共振频率、基频范围和发音方式76% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/716309API
curl https://kongchang.com/api/v1/knowledge/claims/716309MCP
get_claim(id=716309)