Sesem_ Ag

@OfficialLoganK @zapier Sure it outperforms opus 4.7 and gpt

Gemini 3.5 Flash ranks #1 on Automation Bench (from Zapier), beating every other frontier model at a much lower cost

@TerminalWarGame Pubg knock down animation lol

We're testing Terminal War internally. We believe early feedback is valuable but there is much to improve already. Would you be interested in a playtest this simple?

@kimmonismus ...Isenberg? Hm... oooh...!

I’ve been invited by Google to attend its annual I/O conference as part of the Builders Program, and I’m incredibly excited. It’s my first time at Google, and this time I brought a camera with me to capture the experience and create a recap video afterward. During the event, I’ll be conducting two fascinating interviews with Google employees, focusing on AI. These will be published at a later date. I’ve already met some amazing people from the community. Here’s to two unforgettable days!

@synthwavedd i can feel "vibecoded" shi just by look. Whats wrong with you, leo?

I've been testing Gemini 3.5 Flash for a little while now, and I'm excited to be able to share one of the outputs that most impressed me! This was 0-shot, no harness, with a single sentence prompt. It outperformed all Claude models, Gemini models (by far), and arguably GPT-5.5 🔥 The issue of laziness that has plagued Gemini models forever has mostly been consigned to history.

@acasualnpc @OfficialLoganK You judging model's quality by a single very short video... bro, the model is not even out yet

@OfficialLoganK Holy shit, SOTA Gemini video model. Seedance is thing of the past. This changes everything 💀

Gemini

@7a7zz @LexnLin Ughhhh uhh but b-but, it will be on the first place of all benchmarks! Ooga booga boga

@LexnLin why so obsessed with gemini mate , they always disappoint 😭

YES WE WILL GET 3.5 (cuz gem is half of gemini)

@kimmonismus little codemaxxed thing, unusable in real tasks

I love GPT-5.5. It's a workhorse and exactly the model I was hoping for. But the fact that rumors say version 5.6 is already in the starting blocks makes me even more excited! OpenAI is on fire.

@Fabiobuilds @synthwavedd @arena man be real, nobody using flash models or even gemma if they have access to at least 3.1 pro. Not even talking about other models like claude or gpt

@synthwavedd @arena Google is already doing excellent work with small and local models, I think they are not prioritizing right now pro models. Theu seem to have left the game all to OpenAI and Anthropic for that (which are doing quite bad in small / local models)

so apparently gemini 3.2 pro is being tested under "gemini-3.1-pro" on @arena's Code Arena (they have done this kind of stealth testing before) ...and if this is really 3.2 pro, it's not looking good. somehow they gpt-ified frontend? hopefully this is an arena-specific quirk

@Tate_Dollar @Hesamation it's unusable for now, it's just architecture, not a ready project

@Hesamation Why wait for them to release when somebody already coded it themselves? github.com/Jaravus/opensu…

> 12M context window (read it again) > 52x faster than FlashAttention > beats Opus 4.6 on SWE-Bench > 5% the cost of Opus BUT WAIT A MINUTE: > technical blog not technical > access coming soon > paper coming soon > ““Built by researchers from Meta, Google, Oxford, Cambridge, BYU” doesn’t name a single one of them if this is not a scam, or the numbers aren’t dishonest, it’s disgustingly promotional.

@PolymerCan @walkingtoncity @tarkovarena why? Do you have any arguments or you just don't have your own opinion?

@walkingtoncity @tarkovarena Because it’s BSG, they are shit

Gladiators, we invite you to test your knowledge of weapon mods in #TarkovArena. Can you identify all the attachments? Share your answers in the comments!

@chetaslua @tehnlulz acting like it's something amazing and innovative

🚨 GPT 5.5 FIRST OUTPUT FROM CODEX I am super hyped for this launch , btw have you guys heard news about Mythos controversy ( that some forums members had access to it ) credit - @tehnlulz

@flowersslop is this ishowspeed?

Gameplay screenshot of GTA but it is in the Harry Potter world, Hogwarts in the background, wand equipped

@discord ...why

please enjoy

@kimmonismus eh, still no real benefits except for benchshits. Yet another flash level model that loops as fuck

A 12-month time difference between Gemma 3 27b and Gemma 4 31b. The jump is absolutely enormous. Just look at the evaluations between the two models. GPQA doubled, AIME 2026 went from ~20% to ~90%, and so on. Crazy.

@promin_tysson @tomorrowstrand НОГИ

@Klampzy @MaD_MrL Will you watch a movie about the game you like, but every character will look like a shit? No, cuz nobody likes watching a shit, it's just facts

I still don’t get this casting lmao

@ZC2forBO7 @EmpireFaIls @MaD_MrL "Im showing you that i don't care. Please do not offend me" bro:

@EmpireFaIls @MaD_MrL It’s funny that you’re responding as if I actually care about your opinion lmao.

@wholyv smaller model will be more stupid, it literally has less knowledge about everything. Corps trying compensate it with thinking, but, looking at gpshit 5.4, it looks terrible. Just look at claude models.

🚨We all thought Apple was leaving behind in the AI race, turned out they had already crossed the finish line, waiting for the world to catch up. entire world is running behind the concept of bigger model -> better but apple took edge compute seriously and started working on it much earlier, even before openclaw became a thing. We all know how apple is so much obsessed with privacy, it’s no wonder they didn’t want an LLM to get access to your private data. so what’s apple’s plan? Apple’s plan is edge ai. it aims to provide small and efficient and cheap models directly installed on your phone which you can access privately, securely, without uploading your data on any LLM for a fraction of regular cost. won’t small models be stupid? quite the opposite actually. Alibaba launched 4 small models and all of them beat opus 4.5 on many benchmarks. the goal here is targeted use of models, not to make them general purpose. regardless, we will soon see something very cool in the space of AI that will completely redefine how we work with it. this time, it will come from apple.

@4thOfJuly365 Rainbow Dash cosplay??

WTF is this?

@Chief_Kewatan @RespectfulMemes bro made a movie once in a whole century, he deserves a rest

@RespectfulMemes This guy made a whole ass movie on top of the usual YouTube stuff and other business ventures. Mf deserves a rest This post is cringe