Unverified50% confidenceBenchmarkExact time
在实时陪伴场景中，首个token的生成时间（Time to First Token）需要控制在200-500ms以内才能保证体验流畅
1
Sources
50%
Confidence
Long-term
Relevance
8/25/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedIn digital human scenarios, the WebRTC end-to-end latency typically needs to be kept within 500 milliseconds to ensure a natural and smooth interactive experience.78% similarUnverifiedVAD环节需在100-200毫秒内响应，整条链路端到端延迟需控制在1-3秒内才能维持良好对话体验69% similarUnverified对于500 Token的回答，流式输出可以让用户在首Token延迟（约500ms）后就开始看到内容，而非等待完整生成（约10秒）69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/800155API
curl https://kongchang.com/api/v1/knowledge/claims/800155MCP
get_claim(id=800155)