Unverified50% confidenceFactExact time
Two common metrics for evaluating local model speed are time to first token (TTFT) and generation speed (tokens/s).
1
Sources
50%
Confidence
Long-term
Relevance
9/16/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedLatency and throughput are two different technical metrics.74% similarUnverifiedInference latency is typically decomposed into two core metrics: time to first token (TTFT) and subsequent token generation throughput.72% similarUnverified对流式响应场景，应重点关注首Token延迟（TTFT）而非整体完成时间68% similarUnverifiedtokens/s指标可细分为首token延迟（TTFT）和生成速率（Generation Throughput）两个子维度68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/925041API
curl https://kongchang.com/api/v1/knowledge/claims/925041MCP
get_claim(id=925041)