Anonymous
6/20/2025, 11:50:39 PM
No.105655502
>>105655395
I always do slowburns, average of 18.6k input token per request. avg 145 reasoning + 530 output token. Averaging it out over $200 of use I paid $0,72/M for input token.
25k in -> $0,018
2.5k out -> $0,0375
Total $0,0555 per request. But to be honest that 2500 token output value of yours is way too high. I haven't used deepseek in a while but at least the initial R1 used too many reasoning token. With my values from above I ended up at $0,023/request which is fine for me.
I always do slowburns, average of 18.6k input token per request. avg 145 reasoning + 530 output token. Averaging it out over $200 of use I paid $0,72/M for input token.
25k in -> $0,018
2.5k out -> $0,0375
Total $0,0555 per request. But to be honest that 2500 token output value of yours is way too high. I haven't used deepseek in a while but at least the initial R1 used too many reasoning token. With my values from above I ended up at $0,023/request which is fine for me.