Verified80% confidenceFactExact time
Flash模型每token的推理成本通常比Pro级模型低一个数量级
4
Sources
80%
Confidence
Long-term
Relevance
5/22/2026
First Seen
Sources
Gemini 3.5 Flash在GDPval基准上实现巨大飞跃
twitterOfficialLoganK5/22/2026
Related Entities
Related Claims
Unverified轻量版模型单次推理的算力消耗往往只有旗舰版本的十分之一甚至更低70% similarUnverifiedJF透露在95%甚至更高情况下，Flash或Flash-Lite等较小模型作为裁判给出的评分与大模型一致，评判环节往往不需要昂贵的Pro大模型69% similarUnverifiedFlash模型的首Token延迟（TTFT）通常比Pro低50%以上68% similarUnverified千问3.8 Flash Next的稀疏率约为95%，每次推理只动用不到5%的参数66% similarUnverified效率型模型以远低于旗舰模型的推理成本（通常低5-10倍）提供接近旗舰模型80-90%的能力表现66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/40046API
curl https://kongchang.com/api/v1/knowledge/claims/40046MCP
get_claim(id=40046)