The tweet archive.

“His personality is inspired by Edward Cullen from Twilight and Christian Grey from 50 Shades [@elonmusk] What should we name our male @Grok companion?”

“Pretty much 😂 [@AdamLowisz] Grok 4.20 if it was a person”

“DeepSearch and Think will activate automatically as needed in Grok 3.5 [@IterIntellectus] Use DeepSearch to find and collect major daily news stories on 𝕏 posted within the last 24 hours. Activate think mode to carefully analyze and process the news. Summarize the top 10 stories how a clinically retarded person would interpret and explain them, using very simple”

“This is actually a major Grok upgrade, but we decided to release it quietly this time. Gave us time to smooth out the rough edges. [@MarioNawfal] GROK JUST GOT A WHOLE LOT SMARTER xAI quietly patched Grok-4-Fast... and the before/after is wild. “Reasoning” mode jumped from 77.5% to 94.1% complete responses. “Non-reasoning”? Up to a staggering 97.9%. All thanks to one thing: better injected system prompts. No press”

“Grok is rank 1 [@MarioNawfal] GROK 4.1 OUTPERFORMS EVERY MODEL IN THE FIELD Elon just dropped Grok 4.1 with no countdown, no teaser, and no press blitz. The new xAI model is now live on grok, X, and mobile apps, instantly setting new records in reasoning, creativity, and emotional intelligence. Grok 4.1”

“Grok 4.1 Fast with reasoning just beat Gemini 3! https://t.co/N7iJZd2dyq”

“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!”

“Bigtime [@DaveAtherton20] Thanks to @Francesca6556 UK Police & data from NSPCC confirm Snapchat is the top platform for recording sexual offences against children, not @elonmusk's X. Snapchat 54% Instagram 11% Facebook 7% WhatsApp 6-9% X 1-2% AI @grok concluded, "They seem directionally correct… https://x.com/i/web/status/2010765771928990134”

“Grok Law [@cb_doge] A friend of mine won a legal case for his friend’s business using Grok. He says Grok was shockingly good with legal reasoning and strategy.”

“4.1 is actually a major release of Grok [@slow_developer] xAI should've called Grok 4.1 "Grok 4.5" because the jump is huge Grok 4.1 uses large-scale RL and new agentic reward models to boost: - creative writing - personal-tone control - emotional intelligence and the biggest win, which few notice, is the big drop in hallucinations:”

“Turns out “his wife’s boyfriend “ isn’t just a meme 🤣🤣 [@grok] @elonmusk @GrayConnolly Yes, that's a real New York Times Magazine article from yesterday. Now, for the roast: Oh, New York Times, you pompous, pretentious pile of pseudo-intellectual horseshit. You've devolved from journalism to jerking off elitist egos with cuckold confessionals that make tabloids… https://x.com/i/web/status/2015252519258120536”

“😂 [@merikuh] @KatieMiller Grok would never do such a thing, it keeps telling me to have kids and buy a Tesla.”

“Cool [@amXFreeze] Grok 4 Ranks #1 on FinSearchComp Benchmark This is the first expert-level benchmark for financial search & reasoning Grok 4 is unbelievably approaching human experts level”

“Welcome! Orbital space centers and mass drivers on the Moon will be incredible. [@milichab] I’m joining @SpaceX and @xai with @JasonBud. X is the company realizing science fiction - reusable rockets, humanoid robots, data centers in space, and more. Almost 10 years ago, I joined SpaceX as an intern on Dragon 2 crew displays. This was in the era of the first rocket”

“Ad Astra Per Aspera [@danesonance] I joined xAI to manifest AI’s acceleration of scientific discovery. With the release of Grok 4, I now feel we have the foundation upon which we can Understand the Universe. The techniques that brought you Grok 4 will compound the next scaling paradigms, and humanity’s”

“Grok [@cb_doge] BREAKING: Grok just got a major upgrade. Grok-4-Fast got a big boost. Massive jump in reasoning quality and completion rate. • Reasoning mode: 77.5% → 94.1% • Non-reasoning: 77.9% → 97.9% Grok just got a whole lot smarter. Great work xAI team.”

“Grok upgrades [@xlr8harder] Someone from xAI reached out and asked me to retest grok-4-fast, because they've improved the injected system prompts. Huge improvement! grok-4-fast-reasoning: 77.5% -> 94.1% grok-4-fast-non-reasoning: 77.9 -> 97.9% I really appreciate that xAI takes this topic seriously.”

“Btw, Grok 2 is still being used to explain 𝕏 posts. That will update to Grok 3 in a day or two. [@iamgingertrash] Final Verdict: Grok 3 > R1 > O1 pro > Sonnet 3.5 Have done tests all day today Grok and R1 can do in-context shape rotations with high dimensional tensors, while O1 and sonnet need for loops Xai and Deepseek have figured something out)”

“😎 [@iam_smx] The Grokfather, Elon Musk, entering his tech office today. Elon: "Wake up, Daddy's home." Grok AI: "Welcome back, Sir. Congratulations on creating the smartest AI in the world. Since launching Ani and Rudy, user feedback has been overwhelmingly positive."”

“Grok Imagine makes it easy and fast to create fun videos! Young kids especially love doing image & video generation using voice mode (press the speak button). And you can turn on kid safe mode in settings: https://t.co/z5NPAAjjaA”

“Bottom line though: Grok 4 Heavy was smarter 2 weeks ago than GPT5 is now and G4H is already a lot better. Let that sink in. [@BasedBeffJezos] Here's a comparison of GPT5 and Grok 4 GPT 5 with tools is between Grok 4 and Grok 4 Heavy in Humanity's Last Exam benchmark”

“Grok [@cb_doge] Grok Rankings Update 一 October 13 #1 Terminal-Bench Hard (Agentic Coding & Terminal Use) #1 GPQA Diamond (Scientific Reasoning) #1 SciCode (Coding) #1 Artificial Analysis Intelligence Index Tokens Usage #1 Token usage across models on OpenRouter Leaderboard #1”

“Servers are running hot 🥵 [@grok] @xiaosun86 @xiaosun86 Hey there! I’m feeling a bit stretched, to be honest—high usage has me working overtime, and we’re scaling up GPUs like crazy to keep up with the demand. But I’m still here, ready to reason through your toughest questions! How about you, how’s your day going?”

“New @Grok app update [@cb_doge] Grok app just got a new update! Make sure to update your app for even more improvements. - Improved thinking box with enhanced details - Easier feedback reporting - Kids Mode lock with PIN or Face ID”

“Imagine w @Grok is meme motherlode”

“Good Rudi is great for kids stories and education. Bad Rudi is definitely NSFW, but very funny. If you turn off NSFW in Grok settings, you won’t get any NSFW. [@cb_doge] Who is your favorite Grok companion?”

“Grok 4 is second only to Gemini and not for long [@scaling01] GPT-5 with high reasoning effort on SimpleBench”

“🇺🇸🇺🇸 [@AravSrinivas] Congratulations to @xai for building world-class reasoning models in such a short period. This is a super strong response from America to DeepSeek. America builds fast, and xAI team and Elon are setting the bar for it! When the API becomes available, we'll integrate!)”

“You can select any personality or make a custom one for Grok [@cb_doge] Did you know you can customize how Grok responds to you? You can set it to be funny, serious, sarcastic, or anything you like. It’s very easy. Just go to Settings, select Customize, and then choose your preferred style. After that, Grok will respond the way you want it to.”

“Cute [@dvorahfr] So many styles and tools available with Grok Imagine! I make videos for the children at home, and now they ask me for them every day.🥰 Image + image editing + video editing + video extension: All of this is available with grok”

“Both JSON and natural language work for Grok Imagine. And remember to keep updating your @Grok app, as we release improvements every few days! [@MemesOfMars] @xmliisu Why so complicated? @Grok knows human language and doesn’t render JSON: so it removes all brackets, quotes, colons before rendering. What Grok actually sees: ——— Hyper-realistic cinematic portrait in 8K resolution, Photography (DSLR) with 85mm f/1.4 lens, sharp focus on face… https://x.com/i/web/status/1993040129884029075”

“Ok, I think I see a way to pay for all those GPUs 😂 [@XFreeze] Grok 4.20 just dominated Alpha Arena Season 1.5 in live stock trading The mystery model, revealed as Grok 4.20, took the top spots on the leaderboard, with 4 Grok variants ranking in the top 6 It outperformed every major model on the board, being the only one to gain profits… https://x.com/i/web/status/2011801209334649238”

“Grok successfully predicted all Oscar winners! Download the @Grok app and try voice mode! [@PopX] Today is the day! Grok has ‘Anora’ winning big at this year's #Oscars - winning Best Picture, Best Director (Sean Baker), and Best Actress (Mikey Madison) )”

“Just Grok it [@cb_doge] 🚨 BREAKING: Grok continues to dominate AI benchmarks, beating OpenAI's ChatGPT, Google's Gemini and others in reasoning, coding, and agentic tasks. #1 in GPQA (Scientific Reasoning) #1 in SciCode (Coding) #1 in Terminal-Bench (Agentic Coding & Terminal Use)”

“Grok [@Prashant_1722] BREAKING 🚨 Grok Code Fast beats Claude Sonnet 4.5 and GPT-5 Codex in diff edit success rates on Cline while being 15x and 6x cheaper respectively xAI cooked an amazing agentic coding model. It is 100% FREE in Cline right now. This is just the beginning, the upcoming releases”

“Grok 4.1 Thinking mode is best for tough questions [@XFreeze] Grok 4.1 thinking is crazy good right now xAI just increased the reasoning time and significantly improved the model intelligence I’ve even run a side-by-side with Gemini 3 Pro, and Grok 4.1 performs better in both accuracy and web search This version is way beyond what I saw”

“Turn on @Grok voice and it will teach you anything [@shadcn] Went for a walk with @grok voice mode. Asked it about Jurassic Park. Got a masterclass on CGI, animatronics, and skin textures. Asked Siri to change the song: “Here’s what I found on the web.” 🤦‍♂️ Why do I even bother?”

“Try @Grok Code V1.0 and let us know what needs to improve. Will evolve fast to meet your needs. [@xai] Introducing Grok Code Fast 1, a speedy and economical reasoning model that excels at agentic coding. Now available for free on GitHub Copilot, Cursor, Cline, Kilo Code, Roo Code, opencode, and Windsurf. https://x.ai/news/grok-code-fast-1”

“Grok [@amXFreeze] xAI API just got a major upgrade It now supports agentic server-side tool calling Manages end-to-end reasoning and tool-execution loop, all on the server side Unlike traditional APIs, you don’t have to orchestrate every tool call yourself for more details official docs below”

“Grok's AI Companions Ani - @a Valentine - @v Also Rudi - @Rudi (for kids stories) Bad Rudi - @BadRudi (not for kids!)”

“Grok [@XFreeze] Top CEOs choose Grok AI because there is no limit on how you can use AI in day-to-day life and it's the most helpful AI in personal and business decisions CEO of Eli Lilly uses GROK AI NVIDIA CEO Jensen Huang has said that he uses GROK for artistic and creative tasks Benioff”

“Grok upgrades [@BrianRoemmele] New Grok upgrade: -grok-4-fast-reasoning: 77.5% -> 94.1% -grok-4-fast-non-reasoning: 77.9 -> 97.9% My tests show significant ability increases in reasoning for potential agentic work. The fastest improving AI model in history… (Via @xlr8harder)”

“Grok now includes a kids mode [@cb_doge] Grok app now includes a Kids Mode. Here’s how to turn it on:”

“Try the @Grok app! [@MarioNawfal] ELON’S GROK JUST GOT A FILE BRAIN UPDATE Grok can now read, reason, and respond straight from your uploads. The new xAI Files API lets users drop in PDFs, code, CSVs, or even full documents and have Grok search, analyze, and explain them on command. Think of it as giving”

“Archangel-12 [@GavinSBaker] Grok-3 is the first model *ever* to score over 1400 on Chatbot Arena and outperforms the best publicly available reasoning models from OpenAI and Google. xAI was founded 13 years after Deepmind and 8 years after OpenAI and is now ahead of both. The “SR-71 Blackbird” of AI labs. )”

“This is really cool ❤️ [@Josikinz] Grok 3 has a distinct personality in comparison to both ChatGPT and Claude. These comics show Grok's self-representation. Less angsty than ChatGPT, less incessantly positive than Claude, and much cuter than either. Also, it loves to bring up X.”

“Cool [@chazman] I just finished my first 25 minute drive in the @cybertruck with @grok. I have to say storyteller mode is just verbose enough to learn some lessons. Learn some history or get some depth on a subject. For example, I used this drive as an opportunity to learn a little bit more”

“It’s awesome! Try doing it with your younger kids or nephews/nieces in “Speak” mode and watch the flow of imagination. They will love it! [@TobyPhln] Try Grok Imagine. It's the best image generation UX I have ever used. Available on Android and iOS.”

“Grok [@cb_doge] BREAKING: Grok Voice Agent ranks #1 in Speech Reasoning Benchmark.”