Mixed: zhipu/glm-5-3-flash on inferencenet

FieldBeforeAfterChange
Input0.04900.12+144.9%
Output0.500.39-22.4%
Cache read0.01600.0320+100.0%

What this means by workload

Illustrative monthly cost of the changed field(s) alone, at a representative 10M tokens/month split per workload preset. Not a substitute for the calculator, which uses your own volume and every price field.

WorkloadBefore / monthAfter / month
Coding agent$0.72$1.33
RAG chat$1.17$1.60
Summarisation (batch)$0.94$1.47
Classification$0.58$1.25
Chat assistant$2.29$2.27
Content generation$4.10$3.34

Back to model page | All changes