input tokens is not the smaller amount, you’ll use way more input tokens than output, so the cost of input tokens is what you should mostly pay attention to
2026-04-25 22:52:35
1
⚡︎AıPrıest︎⚡︎ :
Additional skils used?
2026-04-26 06:29:35
1
johnoruttley :
V4 would stack well through api ones Hermes’ or open claw then?
2026-06-17 23:42:23
0
Adam Luz :
I have gone back to Claude 4.6 and both 4.7 and 4.6 always down. 4.7 is always telling me to do things myself and is way to cautious. it's getting worse. also saw no upgrade from 4.5 and 4.6. Are influencers using this stuff... it's not crazy different.
2026-04-27 03:29:44
0
Phinithi | AnurIQ 🇿🇦 :
I have a feeling these platforms are capable of near perfect logic. The day debugging become obsolete, it would erode "shareholder value", because ROI is predicated on tokenization. Now I see why DeepSeek, regardless of its benchmark, is a thorn in the west. Oh, that it's not depended on Nvidia should fly in the face of geopolitical strategists.
2026-05-02 12:35:34
0
jeronimovm :
almost no bench marks in reasoning is quite suspicious
2026-05-01 13:48:39
0
Car mc car car :
Codex bad
2026-04-25 19:06:10
0
T :
There isn't a link in the description
2026-04-26 01:03:42
0
Trader's Friend :
🥰🥰🥰
2026-04-29 03:19:51
0
To see more videos from user @chase_ai_, please go to the Tikwm
homepage.