LuoFengCheng (@multuslove)Cursor额度翻倍后,剩余40%的第一方计费池额度的使用量 中发帖

前几天Cursor给之前的Auto+Composer计费额度翻倍了,之前蹬完的pro账户的auto计费池池还剩40%,3天蹬完了。 
老马的4.5还是很好用的,指哪打哪。速度还快,体感比deepseek的flash快。
[图片]




模型
Cache Read
Cache Write
Input
Output
Total
可读输入
缓存读取占比
缓存命中率




Grok
294,860,800
0
19,421,904
2,435,324
316,718,028
314,282,704
93.10%
93.82%



40%的用量大概有3.1亿的token
实际上的Grok的额度应该更高,我是混着composer2.5用的。
总用量




模型
Cache Read
Cache Write
Input
Output
Total
可读输入
缓存读取占比
缓存命中率


...