@TimJayas

Playing with different AI Tools $1M ARR before AGI šŸŽļøšŸ’Ø

Joined August 2025
  • Tweets18.6K
  • Following483
  • Followers5.6K
1.4KMedia
if tibo doesnā€™t ship any big update for codex in this 28 days then itā€™s the end for openai
34
7
610
171,655
kimi-k3.1 pro is better than opus 5.5
šŸ¤ Paid partnership
1
1
42
after testing i can confirm kimi-k3.1 pro is the most bechmaxxed model of 2026 but itā€™s a good model, just not as good as shown in AA benchmarks with a worse usage limits the closest most accurate benchmark iā€™ve found is from arena ai
3
2
16
653
no matter how good kimi-k3.1 pro maybe most people will still prefer deepseek-v4.1-flash-1017 over any other open models
2
1
3
397
im tired of waiting for fable 5.5 just give me opus 5.6 instead
8
1
15
1,024
muse-wonder-full is top 2 model in AA benchmarks yet i rarely see anyone talk about it. heavy underrated or bechmaxxed?
Tim Jayas
@TimJayas6h
itā€™s sad to see gpt models are not even in the top 3 in AA benchmark the new top three ai labs are #1 anthropic #2 meta #3 moonshot does this counts as openaiā€™s downfall?
3
8
544
opus 5.5 was the only last model which was not bechmaxxed and maybe gemini 4 argon too
3
1
30
2,196
itā€™s sad to see gpt models are not even in the top 3 in AA benchmark the new top three ai labs are #1 anthropic #2 meta #3 moonshot does this counts as openaiā€™s downfall?
18
1
29
2,082
I canā€™t even post a 3d three.js output of kimi-k3.1 pro without someone accusing it to be paid post lmao. itā€™s looking rough for kimi
6
17
1,429
i doubt thereā€™s anyone left who still uses gpt-6 astra
28
1
43
3,184
i have a feeling even google and openai does this but I canā€™t prove it Iā€™ve seen creators getting free ultra sub just to make 3d gemini 4 argon posts
Elshayib
@elshayib_23h
That's how Kimi AI are promoting their new model Kimi K3.1 pro and I know all the influencers working with them and none of them are honest,I have posted about it myself but I didn't join this and never used their templates that they provide. PROPAGANDA, I have more to share it's f**cked up I'm gonna have follow up for this
19
2
29
2,035
how can kimi-k3.1 pro be a distill of GPT and yet be better than Astra. that doesnā€™t make sense
10
1
17
1,866
why are so many anthropic employees leaving just weeks before their biggest IPO in history
12
1
32
2,224
bro left anthropic the moment opus 5.5 got nerfed šŸ˜­
Thariq
@trq21214h
Replying to @lydiahallie
wait is it just me or is Opus 5.5 dumber
3
13
971
if i find anyone saying kimi-k3.1 pro is better than opus 5.5 again i'll crash out
10
1
27
1,346
since opus 5.5 is nerfed is it now time to switch to codex?
22
3
38
3,610
no way kimi-k3.1 pro is already being accused to be distill of gpt-6 astra
2
4
489
so the only reason openai isn't launching a model significantly better than opus 5.5 is because it's too powerful? openai's internal models are so strong they had to cut off the internet access to stop them from escaping the sandbox and since even this wasn't enough they had to launch a high compute red-teaming campaign that uses frontier models to find vulnerabilities
2
1
452
openai.com/index/securing-ouā€¦

Securing our frontier model training environment

OpenAI explains how we strengthened containment for frontier model training and evaluation, including offline environments and high-compute red-teaming.

openai.com
102
i hate to say this but openai just lost the race. china just released kimi-k3.1 pro which outperforms gpt-6.1 sol and astra and even matches opus 5.5.. idk if even 6.1 astra could save openai now
9
29
1,523
opus 5.5 build an entire 3d office using three.js and added me as the head of partnerships šŸ˜­
Jesse Rowe
@jessetroweOct 23
I asked Claude to build an office for the @tavus PALs... I was not expecting it to build THAT kind of office
2
15
875
you can even talk to me lmao here: office.tavus.io

The PAL's Office Ā· Tavus

Walk in, find a desk, press E. The PALs see you, hear you and talk back.

office.tavus.io
90
Load more