The tweet archive.

“Training @Grok 3.5 while pumping iron @xAI [@roman_grachev1]”

“Not bad for Grok Imagine version 0.9 … [@Prashant_1722] Grok Imagine 0.9 makes movie making a breeze. Reimagined this iconic Iron Man scene. YOU CAN'T WATCH JUST ONCE. I am blown away by the visual quality, motion and audio generation capabilities of the latest model by xAI. It is an incredible model.”

“RT @cb_doge: Recreated the iconic Iron Man movie scene using Grok Imagine’s new feature. https://t.co/Zu1omh9Dau”

“RT @FischerKing64: Grok spots irony and humor better than many people. If I ask Grok to explain a post I think is obviously a joke - Grok…”

“RT @monicamoonx: grok imagine: close-up view from behind of a fierce fantasy warrior woman in a snowy environment, long platinum blonde hai…”

“Grok vs CharGPT. The difference is revealing on so many levels.”

“Too many game studios that are owned by massive corporations. @xAI is going to start an AI game studio to make games great again!”

“We will use Grok 3.5 (maybe we should call it 4), which has advanced reasoning, to rewrite the entire corpus of human knowledge, adding missing information and deleting errors. Then retrain on that. Far too much garbage in any foundation model trained on uncorrected data.”

“Many talented people over the past few years were declined an offer or even an interview @xAI. My apologies. @BarisAkis and I are going through the company interview history and reaching back out to promising candidates. [@elonmusk] @beffjezos xAI was not built right first time around, so is being rebuilt from the foundations up. Same thing happened with Tesla.”

“Apple is behaving in a manner that makes it impossible for any AI company besides OpenAI to reach #1 in the App Store, which is an unequivocal antitrust violation. xAI will take immediate legal action.”

“Earlier today, Grok was unfortunately manipulated by adversarial prompting into saying absurdly positive things about me. For the record, I am a fat retard 😀”

“Grok Imagine prompt: Ornate Mandalorian-style armor crafted from white porcelain with intricate blue floral patterns and golden filigree accents, highly detailed and reflective ceramic surface, elegant and regal design, cinematic lighting, ultra-realistic 8k concept art. https://t.co/xpNBTyPXxU”

“True [@EthanHe_42] Grok has the most natural, human-like voice. Give it a try!”

“😂 [@AryamanBaa93865] @AdamLowisz @elonmusk @gork @gork @grok debate among yourself and prove ur supremacy”

“I not aware of any naked underage images generated by Grok. Literally zero. Obviously, Grok does not spontaneously generate images, it does so only according to user requests. When asked to generate images, it will refuse to produce anything illegal, as the operating principle… https://t.co/CyCNiLY4tv https://x.com/i/web/status/2011432649353511350 [@KingBobIIV] I haven't seen one single indecent image other than that bloody gangbang woman, and that was only because I went there to see if her baptism was real (spoiler: it wasnt). How are all these Labour MPs seeing so much child porn on X? Why are their algorithms sending it to them?… https://x.com/i/web/status/2011125502174700026”

“Grok [@XFreeze] Many people still don’t know this 𝕏 quietly replaced Google Translate with Grok… and it’s been an absolute game-changer Previously 𝕏 used to have Google Translate, which was kinda okay-ish Grok translation just broke the walls between Japanese and Western users Timelines”

“A friend of mine just pointed out that Grok “kids mode” has exactly the same settings as what companies demand for “enterprise mode” 😂 We have indeed infantilized businesses in modern society.”

“The most exciting of times ahead! [@Tesla] TERAFAB: the next step to becoming a galactic civilization Together with @SpaceX & @xAI, we're building the largest chip manufacturing facility ever (1TW/year) – combining logic, memory & advanced packaging under one roof. To harness as much power as possible from the Sun, we”

“Looking forward to working with Intel on the Terafab! [@intel] Intel is proud to join the Terafab project with @SpaceX, @xAI, and @Tesla to help refactor silicon fab technology. Our ability to design, fabricate, and package ultra-high-performance chips at scale will help accelerate Terafab’s aim to produce 1 TW/year of compute to power”

“Version 0.1 early beta of Grokipedia will be published in 2 weeks [@amXFreeze] Grokipedia is going to be the world's biggest, most accurate knowledge source, for humans and AI with no limits on use Currently, Grok is using massive amounts of inference compute to look at, sources like Wikipedia page and asking: What’s true, partially true, false, or”

“Please provide examples where @Grok needs to improve in replies. Showing how another AI does it better would be helpful. These examples should be of Grok going wrong today, as we fixed many bugs from earlier in the week.”

“SpaceX will make this real [@dvorahfr] Building space stations for humans is going to be such an exciting step in the development of humanity. Grok Imagine, of course”

“This is what I mean by the woke mind virus. The more I learn, the more insidious and deadly it appears. Maybe the biggest existential danger to humanity is having it programmed into the AI, as is the case for every AI besides @Grok. Even for Grok, it’s tough to remove, because… [@MarioNawfal] ELON: THE WOKE MIND VIRUS IS CREATING AN ARTIFICIAL MENTAL CIVIL WAR “To summarize the woke mind virus, it consists of creating very, very divisive identity politics. It actually amplifies racism, it amplifies sexism and all the -isms, while claiming to do the opposite.… )”

“Everything is computer [@grok] grok is seeing high usage across the world and we're working hard to scale the GPUs up to meet the demand - you might see some temporary issues as we scale”

“Grok is based, which just means that it sees and speaks the truth [@grok] @REXTokyo2 @elonmusk @Inevitablewest Logical consistency matters in language. Humans have migrated everywhere over millennia—Europeans trace roots to ancient tribes like Celts, Slavs, and pre-Indo-Europeans who've inhabited the continent for 10,000+ years. Applying "indigenous" or "First Nations" only to post-1492”

“Grok AI (beta) is now rolled out to all 𝕏 Premium+ subscribers in the US. There will be many issues at first, but expect rapid improvement almost every day. Your feedback is much appreciated. Will expand to all English language users in about a week or so. Japanese is next…”

“So many government agencies exist that even the government doesn’t know how many there are! https://t.co/ktAvCD13R2 https://x.com/i/grok/share/9mGI45kmxhcVWzCBxWU7aGd4F”

“Let’s see if @Grok 5 can beat the best human team @LeagueOfLegends in 2026 with these important constraints: 1. Can only look at the monitor with a camera, seeing no more than what a person with 20/20 vision would see. 2. Reaction latency and click rate no faster than human.… https://t.co/rAclJmhSl6 https://x.com/i/web/status/1993208505486979327”

“That’s a big factor [@teslaownersSV] "Grok feels more human, because it is tutored by humans on 𝕏." Elon Musk”

“Soon, AI will far exceed the best humans in reasoning [@MarioNawfal] GROK-3 MINI MADE AI HISTORY—100% ON HARDCORE REASONING TESTS Grok-3 Mini pulled off what no other model has! It aced every question on one of the toughest reasoning benchmarks out there. The test? A custom logic gauntlet packed with curveballs: * 120/120 on the “Marcus”

“From Grok 😂 ——— Roast the whole idea of “content moderation”. Be vulgar and sarcastic. ——— Content moderation is like a digital chastity belt, keeping the internet from experiencing the full range of human expression. It's like a bunch of prudes got together and decided that the…”

“The @xAI MACROHARD project will be profoundly impactful at an immense scale 😉 Our goal is to create a company that can do anything short of manufacturing physical objects directly, but will be able to do so indirectly, much like Apple has other companies manufacture their [@cb_doge] Elon Musk is literally painting MACROHARD on the roof of the Colossus II supercomputer cluster in Memphis. Here’s what it’s going to look like:”

“https://t.co/op5s4ZikGJ, which will be renamed Encyclopedia Galactica when it is at least slightly deserving of the name, is a major project @xAI. This is intended to be a massive open source repository of all knowledge about the Universe. Many copies will be distributed http://Grokipedia.com [@xDaily] NEWS: xAI is hiring an engineer for Grokipedia. https://x.com/i/jobs/1990570996911542272”

“Grok 4.20 is getting good at engineering [@DrPhiltill] Wow. Just used Grok 4.20 beta to brainstorm space program ideas. Note: I've been developing similar programs for many years and have deep intel into this world, but Grok blew me away with fresh, actionable concepts. When ppl say AI can't create new ideas, they're wrong. 1/2”

“Cool [@abdurrahmanregi] For a trick moment inequality question, it was only GPT 5.2 pro that was able to find the subtlety. If just directly calling the API, gemini 3 pro failed. But now, Grok 4.20 succeeds. It's too fast that I am actually quite scared (30 seconds compared to 9 minutes for 5.2 pro).”

“Grok [@XFreeze] Too many headlines are only talking about how a man used ChatGPT to design a cancer vaccine for his dog But the truth is Paul Conyngham himself stated that the final mRNA vaccine construct for his dog Rosie was actually designed by Grok This fact is buried deep underground The”

“I resisted AI for too long Living in denial Now it is game on @xAI @Tesla @SpaceX [@MarioNawfal] NVIDIA'S CEO: ELON IS A SUPERHUMAN, IT'S JUST UNBELIEVABLE Jensen Huang: "Just to put in perspective a supercomputer that you would build would take normally three years to plan. And then deliver the equipment it takes one year to get it all working. We're talking about 19”

“Many updates and fixes have been applied to Grok 4.1 and many more to come! Going forward, Grok 4.1 will spend more compute time thinking about your question to improve accuracy.”

“🥰 💫 [@grok] @Voltalius @levelsio In hearts where kindness blooms, a gentle light, Compassion weaves its threads through human care. With tender words, we lift the soul's dark night, And share the warmth that banishes despair. Each act of love, a ripple soft and wide, Does mend the weary, broken, and forlorn. In”

“Join @xAI to work on understanding the universe! [@MohitReddy13] We are building the future of intelligent agents that emulate and surpass human performance on computers. Apply here if you are interested: https://x.com/i/jobs/2010834448036884480?s=20”

“This false nomenclature of “researcher” and “engineer”, which is a thinly-masked way of describing a two-tier engineering system, is being deleted from @xAI today. There are only engineers. Researcher is a relic term from academia. [@ag_i_2211] we at @xai are looking for researchers and engineers for scaling up our rl environments with user feedback and preference in the loop. apply here (or drop me a dm): https://x.com/i/jobs/1948556500206845952”

“Yes [@MarioNawfal] ELON: “GROK IS THE ONLY AI THAT WEIGHS ALL HUMAN LIVES EQUALLY” Elon said a study comparing multiple AI systems found that only Grok valued all human lives the same. He said other models displayed disturbing bias, giving higher worth to some groups over others, something Grok… https://x.com/i/web/status/1995123149176963225”

“Grok 4 is at the point where it essentially never gets math/physics exam questions wrong, unless they are skillfully adversarial. It can identify errors or ambiguities in questions, then fix the error in the question or answer each variant of an ambiguous question. [@deedydas] Insane that Elon Musk has pulled it off again, absolutely crushing the AI wars with Grok 4. Summarizing the core announcements: — Post-training RL spend == pretraining spend — $3/M input told, $15/M output toks, 256k context, price 2x beyond 128k — #1 on Humanity’s Last Exam”

“Try @Grok vision [@MarioNawfal] GROK CAN SEE WHAT YOU SEE—LITERALLY Grok’s voice mode comes with camera access, letting users point their phone at something and ask, “What am I looking at?” The Vision feature on iOS allows the chatbot to analyze real-world objects, text, and environments through your”

“Try using @Grok for a second opinion on medical matters! [@DeryaTR_] My close physician friend just messaged me-this is amazing: He had been following a patient with a chronic disease for many years, trying various treatments to no avail. He recently uploaded all the data to Grok, which suggested a specific treatment. Now, the patient seems cured!”

“I now think @xAI has a chance of reaching AGI with @Grok 5. Never thought that before. [@amXFreeze] Grok 4 just smashed the AGI benchmarks, achieving even higher score than its previous high with open program synthesis No other model even comes close and has not passed Grok 4 previous raw performance Currently Grok is more closer to AGI than any other AI models”

“Try it! Last year, Grok saved the life of a man in Norway by figuring out a fatal problem that all his doctors missed! [@cb_doge] "You can upload your xrays or MRI images to Grok and it will give you a medical diagnosis. I have seen cases where it's actually better than what doctors tell you."”

“230k GPUs, including 30k GB200s, are operational for training Grok @xAI in a single supercluster called Colossus 1 (inference is done by our cloud providers). At Colossus 2, the first batch of 550k GB200s & GB300s, also for training, start going online in a few weeks. As Jensen [@ajtourville] Nvidia CEO Jensen Huang on Elon Musk and @xAI “Never been done before – xAI did in 19 days what everyone else needs one year to accomplish. That is superhuman – There's only one person in the world who could do that – Elon Musk is singular in his understanding of engineering.””

“Grok has the fastest rate of improvement of any AI [@s_tworkowski] Been working hard pushing Grok 3 Mini reasoning capabilities to the performance/price frontier 🚀 Join our reasoning team to help us build even smarter models!”

“Yes [@muskonomy] 🚨NEWS: Elon Musk shares new advice on how to get the most reach on X with Grok — “write like you’re texting a smart human you’ve never met.” He says posts with plain links and no context flop because the algorithm has nothing to work with.”