Unverified50% confidenceBenchmarkExact time
Flash模型的首Token延迟（TTFT）通常比Pro低50%以上
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
8/12/2026
First Seen
Valid until: 11/10/2026
Sources
Related Claims
UnverifiedThe per-token inference cost of Flash models is typically one order of magnitude lower than that of Pro-level models.68% similarUnverified一个70亿参数的模型在RTX 4090上的推理延迟约为50-200ms57% similarUnverifiedAccording to shared data, the Shen Xiao I system reduced the time needed to handle matters by more than half, with an answer accuracy rate exceeding 95%.56% similarUnverifiedThe L1 vector router achieves an accuracy of over 95% on the top-20 high-frequency intents, only 1.5 percentage points below the L2 large model, but with approximately 30× lower latency and cost.55% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/733625API
curl https://kongchang.com/api/v1/knowledge/claims/733625MCP
get_claim(id=733625)