Ashen
Region: US
Monday 21 September 2026 17:39:35 GMT
6431
204
29
5
Music
Download
Comments
Warcrimes :
grok 4.7 seems better at coding but it's terrible on token burn
2026-09-22 12:47:23
0
…… :
chat gpt be like:
2026-09-22 04:14:14
0
Michaello100 :
Elon kept hyping it saying it was better than fable. Remember this guy is calling sam Scam Altman
2026-09-21 17:44:58
8
nacho :
well that only one benchmark you showed. the artifical anlysis score is just below 5.6 sol
2026-09-21 18:08:29
6
Muhammad Gudaji :
ain't nobody using grok LMAOO apart from devs who use cursor, and I don't think any real dev doing serious work is gonna use cursor lol
2026-09-21 18:25:21
4
imaginaryman :
Why do I think the world would be better if google, meta, nvidia and elon all got together and did a concerted hard core focus in on an open source set of models to do better? No one needs 100 crappy models. Heck, you could even start from a better ranked open model and add more goodness, if you can’t do better otherwise. 😳
2026-09-21 21:03:02
0
✝️MasterManea✝️ :
Yuh makes sense I have to fight grok bot for a lot of things I fix it then it makes the same mistake
2026-09-22 08:51:15
0
Shane Twilo :
I don’t like grok, but that source is very unreliable
2026-09-21 21:31:32
0
Rob Hartman :
I personally use Grok for non politically correct views and info.
2026-09-21 22:45:46
0
The AI Paragon :
I believe that they misrepresent the data. My second question is, do you work, collaborate, or are involved with the company "Artificial Analysis" in any form or manner?
2026-09-21 23:19:54
0
Ronny :
I’m interested to see if the grok community will protest for a benchmark update on artificial analysis like the codex community did for for Astra’s initial rankings which ended up changing its score😅
2026-09-21 17:58:29
1
Jorge Ramos :
gotta do a stream showing it off!
2026-09-21 18:33:43
1
ratladder.com :
Ashen do the give away already man
2026-09-21 18:55:05
0
KZM8 :
Yeah it’s because their benchmarks is ran with their own special ”harness”.
2026-09-21 23:11:27
0
salvy :
that benchmark is dead tho not acurate at all
2026-09-21 18:45:34
1
🗺️ :
if the world's richest man wants numbers that say it's the best...well why not?
2026-09-22 01:26:28
0
MQ :
I think grok it will be same is Gemini 4 in some time
2026-09-21 18:16:12
1
lohanmangaka262 :
No bro artificial analysis is a shill benchmark you can literally pay the guys to rank you higher that's what meta did for 1.3 muse
2026-09-22 02:55:14
0
Richst :
Do you even understand the intelligence analysis? It is an average it does not determine the general intelligence capabilities.😂😂😂 so if you are saying they’re lying, you have to be specific they’re lying on what benchmarks and they always show the exact benchmarks in which they win compared to the other models .
2026-09-21 19:43:22
0
Sana Style :
👍
2026-09-22 04:43:35
0
To see more videos from user @ashenbot, please go to the Tikwm
homepage.
