Hacker Timesnew | past | comments | ask | show | jobs | submitlogin

The 2600 tok/s is an "aggregate", not the actual throughput.


yes it is 213 tok/s single stream (so per user)


that 213 wasn't achieved when saturated though. was probably more like 30 tps per stream when doing 2.6k tps.


So per subagent*.


*per stream, I guess is more accurate than either?




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: