Alex Intelligence
Region: CA
Tuesday 25 November 2025 04:25:36 GMT
79022
3435
97
191
Music
Download
Comments
saleaway00 :
no one thinks openAI model is still the best model
2025-11-25 17:59:03
180
mo :
gemini 2.5, 3 and claude sonnet/opus 4.5 are the best models Imo, chat gpt is completely lobotomized and hallucinates so damn much, while the gemini models are so much better at grounding themselves
2025-11-25 22:32:58
73
404 Not Found :
which LLM would you suggest for coding smartcontracts and code review?
2025-11-26 11:21:44
0
عمر :
Agree except the last part anthropic’s claude is better overall
2025-11-25 05:12:58
60
Dreco :
I don’t care about benchmarks, I just know that when I have a difficult task that deepseek and chatgpt fail multiple times and I give it to Gemini, it solves it
2025-11-25 13:31:36
21
gdew21 :
lol open AI does not have the best model
2025-11-25 13:01:33
48
Sypolar :
Do you even use AI? Claude had the best model even before Opus 4.5. Codex is not close to being better than Claude.
2025-11-25 12:44:17
3
ak :
ChatGPT has the best model? Opus???
2025-11-25 11:58:41
6
DLG :
On my private eval set, Gemini 3 is significantly better than GPT 5.1. I don’t love google but gotta give it to them. About to test out new Opus…
2025-11-25 15:02:40
4
abdalazeez :
at last i found someone who knows that benchmarks are easily gamed and are useless..lucky me😁...the only genuine LLM that is worth considering is claude, by far....
2025-11-25 14:55:06
4
no mercy :
I was sceptical about gemini 3, because 2.5 is pretty awful in every way. But 3rd is not bad at all! Especially in coding, its 100% not worse then gpt and claude sonnet overall
2025-11-25 23:42:10
0
Solo Game Developer :
Thay too aggressive with Gemini… it’s always in depression. When it make the errors it say sorry sorry than panic and do not fix anything 😭 I love Claude. It said something: OMG 😱 you right 😂 and than calmly recover or fix error if there is any.
2025-11-25 09:31:14
4
sycicnok :
lol this guy is funny
2025-11-25 19:45:29
1
jimit :
Oh no no no! ChatGPT has gone really bad since Model 5. Claude is definitely the current best model (as of Nov 2025. Who knows what will happen in December?)
2025-11-25 17:15:23
0
42 :
CodeX is better than Gemini 3 pro at writing code in my opinion.
2025-11-25 18:32:11
1
simsim314 :
its pretty good - it for example downloaded, compiled and introduced modifications I asked in nautililus, linux filebrowser. You can give it any source of any reasonable size, and expect working output. Nothing else comes even near it for 10K lines of code or more.
2025-11-25 13:59:48
0
jordanlister :
Claude definitely is the best model with literally no competition
2025-11-25 21:33:36
0
green013 boyyy :
well idk, I don't think chatgpt has the best model, claude, deepseek, and didn't test it but also the kimi sounds good
2025-11-26 10:56:24
0
Jacob Contreras :
I have had the refuse to acknowledge its 2025 times on Claude models, almost exactly the same as the Twitter post. It didn’t believe me until a web search. This isn’t new behavior with Gemini.
2025-11-25 16:51:35
0
Doggies fan :
I’ve found gpt for architecture and planning, Claude for back end and logic and Gemini for front end is a great combo.
2025-11-26 05:33:56
1
🧠aevion.ai- Ai Truth verified :
Sounds like the space race 😅
2025-11-25 17:06:13
0
Nigel :
this is cope. that wasnt what andrej was arguing
2025-11-26 07:21:33
1
λ :
I feel that Gemini gives me better answers. But it’s a lot slower gpt 5
2025-11-26 10:14:54
0
AI & Automations :
nice one 👌🔥
2025-11-25 19:49:18
0
bashsjss :
please talk about sketchy Claude Opus 4.5 loophole discovery in the t²-bench,airline domain
2025-11-26 00:20:54
0
To see more videos from user @ai_with_alexan, please go to the Tikwm
homepage.
