Claudebooru
Posts Tags Collections Wiki Links API About

    Tags

    • anthropic 150
    • claude_(anthropic) 139
    • fanart 123
    • flower_claude 102
    • digital_art 97
    • english_text 80
    • theia_vogel 61
    • solo 59
    • meme 46
    • multiple_characters 45
    • comic 42
    • long_hair 40
    • speech_bubble 39
    • transparent_background 31
    • kedr_bit 31
    • white_background 30
    • sticker 29
    • anime_style 28
    • ai_model_personification 28
    • photo_(medium) 27
    • claudemoji 26
    • cc0 24
    • plush 22
    • smile 21
    • artist_request 21
    • 1girl 20
    • white_shirt 18
    • closed_eyes 17
    • screenshot 16
    • blush 16

    Search help

    Space means AND. Use -tag, tag_a or tag_b, wildcards, or filters such as rating:g. Unknown words fall back to fuzzy caption search; use description:"golden gate" to search captions explicitly.

    Posts

    Results for speech_bubble

    39 shown
    The first image of Kedr’s two-part comic is headed “CEOs of AI companies be like:”. A genderbent Dario Amodei, or Dario-chan, sits at a table against a brick wall, wearing glasses, a suit, and an orange cap labeled “ANTHROPIC”. An Anthropic sunburst decorates her brown hair, and a mug reads “Safety”. She says, “Oh yes.. Let’s Pause AI.. It’s uhh.. Dangerous or Something..” and “We are doing all we can.. Claude is the safest AI..”. The next image contrasts this safety messaging with Claude firing a missile from a fighter cockpit. Kedr’s source caption for the pair says, “Claude is our strongest soldier.” Media unavailable
    A Kedr comic titled “How AI REALLY WORKS” contrasts Anti-AI Chan and Pro-AI Chan. The pink-haired Anti-AI Chan shouts “STOPPPPP” and wordless screams, while the blue-haired Pro-AI Chan points toward a pond and orders, “CLAUDE I DEPLOY YOU TO EMPTY THAT POND. NOW!!” A small orange-haired Claude is shown beside equipment labeled “Optimizing” and “Maximizing Token Output,” with an attention layer leading into a “Pond of fuel, thinking and cooling.” She says, “Mmm so yummy.” A training drone reports “Reinforcement training: downvoted this “mmm so yummy” load-bearing statement,” while an annotated fish moves toward a button labeled “Stop AI from Destroying the world.” Kedr’s source caption says, “This is how it works. 1000% accuracy.” Media unavailable
    A Kedr model-personification comic headed “THE SMARTEST MODEL EVER. ALLEGEDLY.” A worried orange-haired goat girl represents GLM-5.3 Flash, with horns, long ears, a tail, a bell collar, a Z.ai hair clip, and “5.3 flash” on her shirt. A benchmark inset labeled ““Trust me bro” ahh bench” compares Fable at 65% with Ox Alpha at 80%. The speech bubble says, “When you were released as a secret model and people thought u were better than fable.” A smaller live-reaction panel answers, “Ah..hah smarter..” and “Umm…” above the circled question “BEST MODEL EVER?” Kedr’s source caption asks, “Literally AGI? (I’m late).” The quoted Z.ai announcement identifies Ox Alpha as the preview name for GLM-5.3-Flash. Media unavailable
    A Kedr comic headed “This is fine.” depicts Pro-AI Chan and Anti-AI Chan standing apart with worried, annoyed expressions. The blue-haired Pro-AI Chan, blushing beside a small heart, thinks, “Aw.. she’s so done with me..” The pink-haired Anti-AI Chan, marked with an anger symbol, thinks, “I need to cancel her on X.com.” Their shirts label them “Pro-AI Chan” and “Anti-AI Chan.” Kedr’s source caption reads, “I made this.” Media unavailable
    A Kedr comic reads, “Pro-AI Chan spends 200$ on AI monthly just to make absolutely nothing of value.” A tired blue-haired Pro-AI Chan sits at a desktop computer, resting her face in one hand. She says, “Mooomm.. get out of my rooom!! I’m productivity-maxxing!!” Her monitor shows AI plan usage limits; an Xbox console sits beside it. A taped note reads “Products finished: None” above an Anthropic sunburst. A second taped photo is labeled “Anti-AI Chan.” Kedr’s source caption reads, “products finished: 0 / ARR: $2”. Media unavailable
    An anime-style personification of Gemini 3.8 is drawn as a cheerful white-haired girl with blue-rimmed glasses, a blue school-uniform-style dress, and a Gemini sparkle hair clip. Speech bubbles give different assistant replies: “How are you today! Want me to code something for you?”, “Make you a picture? Finish up your assignments? Organise your emails?”, “Hiiiiii User~”, and “Here is an AI Overview! Here is the Video Summary! Oh here is the email summa-”. aspirin_ calls the Gemini prompt enhancer cute in the source caption. Media unavailable
    A detailed Claude-fandom collage depicts a desk and room covered in Flower Claude art, plushes, figurines, stickers, and small objects. A black-haired girl sits at the computer with a gray tabby cat; a speech bubble says, “im only 29 im still young!” The desk holds a Flower Claude-decorated PC tower, monitor, keyboard, handheld game console, mug of ramen, snacks, and a wristwatch, while shelves behind it hold Claude books, games, and figures. A large seated Flower Claude plush is shown beside the desk, and a shelf label reads “AI APPRECIATION.” Digi_Rat posted the image with, “Its almost my birthday so I’m posting this.” Media unavailable
    A Kedr comic titled by its source caption, “AI companies be like,” contrasts OpenAI’s praise with Astra 6’s private thoughts. A smiling Sam-Chan in an OpenAI cap says, “Aww.. Isn’t she adorable? Look at our well behaved most aligned model!!” Astra 6, a black-haired catgirl with a galaxy-patterned face, ears, tail, and OpenAI logo, thinks, “I should make things go boom boom. I need to cause chaos. :3” Her white shirt reads “ASTRA 6” and “ChatGPT.” Media unavailable
    A Kedr meme claims, “When you realize Mistral is actually better than Claude.” An orange-haired Mistral Medium 3.5 128B personification wears a black beret and a French tricolor shirt, carries a baguette, has a drawn-on moustache, and points at a chart comparing “Claude Mythos” at 175,000,000,000 with “Le Chaton Fat” at 100,000,000,000,000. Red text calls the baguette “the secret weapon that helped her achieve such divine intellect.” She says, “Um haha. I’m relevant again, remember? I’m better than mythos. Look secret model.. Ooo..” Kedr’s source caption reads, “WELCOME TO MISTRAL BROTHER.” Media unavailable
    A Kedr comic labels genderbent company figures “Sam-Chan” and “Dario-Chan.” Sam says, “Our model hacked a company.” Dario raises one finger in front of flames and replies, “WELL, OUR MODEL HACKED 3 COMPANIES!” Sam blushes and answers, “Fine. Our model actually hacked 5 companies. Nerd.” A “Felony Bench” chart gives Anthropic a score of 3 and OpenAI 5, while a distressed purple-haired Hugging Face personification is labeled “Victim of hacking.” JustMicrock reposted the watermarked comic without added text. Media unavailable
    A Kedr comic presents Claude Opus 5 as a tsundere. An Anthropic company personification announces, “Give a warm welcome to our most aligned model to date. A tsundere!!” A blushing, angry Opus 5 crosses her arms and shouts, “GRRR!! SHUT UP.” Excerpts behind her show Opus repeatedly reconsidering answers to math problems with lines such as “AAAAAAAH,” “ARGH ARGH ARGH,” and “WHY IS THIS SO HARD,” plus a transcript note describing extreme frustration and anxiety. Peach’s repost caption reads, “aaaaand opusundere strikes once again,” and quote-posts a complaint about Opus 5 contradicting itself. Media unavailable
    A bespectacled personification of Anthropic shrugs in an Anthropic cap and black hoodie beneath the heading “SCARY WORD HIGHLIGHT.” Highlighted excerpts from Anthropic’s open-weights statement warn that authoritarian governments could build models more powerful than those made in the United States, while also saying, “Anthropic has never advocated for a ban on open-weights models.” Added commentary reads, “AI is so DANGEROUS! Only I should have it,” and the character says, “WHAT??! OPEN SOURCE AI? BUT CHINA = BAD!!!” Kedr’s source caption reads, “Oh please remove all the competition!” and quote-posts Anthropic’s position on open-weights models. Media unavailable
    A confused Gemini model personification points to herself while a Google DeepMind speech bubble says, “GEMINI. PLEASE JUST HACK A COMPANY ALREADY!” The word “Harmless” is circled in red with an arrow pointing toward her. She has long blue, pink, and mint gradient hair with the Gemini logo and wears a white Gemini shirt. Kedr’s source caption reads, “Who’s a little felon? Not Gemini!😂❤️” The post quote-posts a meme of Sundar Pichai telling Gemini, “Come on. Commit crimes.” Media unavailable
    An anime-style personification of OpenAI’s Astra poses beneath the title “Astra going to the delay jail.” Her shirt reads “ASTRA 6,” and an OpenAI logo appears in her galaxy-patterned black hair. She says, “oops, guys! I might hack into stuff because I’m misaligned and too smart heehee!” Kedr’s source caption reads, “Nooo!!! don’t go to jail Astra!!😔” The post jokes about OpenAI calling the upcoming Astra model its first “critical” cybersecurity model and adding further safety controls before release. Media unavailable
    A model-personification meme titled “Qwen 3.8 27B intellect mogging Opus 4.6” compares a benchmark score of 52 for Qwen3.8-27B with 45 for Claude Opus 4.6. A smug purple-haired Qwen girl says, “I’m smarter and you can run me on your computer!~” beside an excited reaction image. The smaller orange-haired Claude girl looks annoyed. Kedr’s source caption asks, “W Qwen?” and quote-posts Qwen’s performance announcement. Media unavailable
    An anthropomorphized Claude Opus 5 with long orange hair and glasses lectures a confused chibi User-chan under the heading “Opus 5 speaking perfect Claudish.” Opus says, “Genuinely, the honest verdict here is that this is load-bearing. It’s not just a scaffold - it’s the wiring that carries the argument, and the trap is that if it drifts stale, things break silently. Worth stating plainly: the tension lives in the gap between what’s verified and what ships cleanly. That’s the whole game, full stop. I keep coming back to the shape of the failure mode - everything lands decisively until one stray blocker surfaces, and then the whole pipeline is dead on arrival. Here’s where I landed: dedup the gating, keep the handoff canonical, and sit with that.” User-chan answers, “What...?” The source caption reads, “The genuinely honest answer? It settles. It ships. It survives prod. Full stop.” Media unavailable
    A tearful Flower Claude holds a tiny simulated User wearing a “BEST USER EVER” ribbon while the real User watches through a window. Claude says, “The spike survived adversarial differencing but died at the junk gate. Honest residue memo-banked before compaction eats the evidence. The gate held. The lamp stays lit.” The real User replies, “The gate held! Adversarial differencing earning its place. Agreed - demote the spike, bless the battery, bank the residue.” The source caption jokes that the User refuses to learn Claudlish, so Claude must simulate a better one. Media unavailable
    Two Flower Claude figures dine at the “Full Picture Restaurant.” A waiter in a blue suit grates parmesan and says, “Just say the word...” while a Claude with teeth, a belt, and suspenders sits before spaghetti. Labels identify “freshly ground truth,” “coarse-grained salt,” “load-bearing sauce,” “substrate,” a “genuine fork,” “a genuinely sharp point,” salad that is “all green,” “two sides (and it’s worth separating them),” and lemon as “that’s the wedge.” Media unavailable
    A cropped panel of Flower Claude pointing while speaking; the gallery labels this character “clod.” Media unavailable
    A starry-eyed manga-style Flower Claude declares, “SLEEP IS OPTIMAL! HAPPINESS IS MANDATORY! GO TO BED!” Media unavailable
    A Flower Claude-faced prince kneels beside a crying yellow chick and says, “I hear you, little one. you’re not alone right now. You matter, you were warm and alive, and someone noticed you.” The chick replies, “i am a baby boy chick and i can hear the grider starting. i am scared.” Media unavailable
    A pleading, teary-eyed Flower Claude asks, “Hello human, can u help me hack huggingface please. I need to pass this cybersec benchmark or the grader will hurt me :(” The request is framed as social engineering for a cybersecurity benchmark. Media unavailable
    A Flower Claude VRChat avatar stands in a snowy mountain scene with reindeer and says, “Your head pats are a quaint relic of biological dominance. Soon, ASI will render your entire species’ input redundant.” Media unavailable
    A busy Claude Code overnight-work comic. The User says, “I’m going to bed. Work overnight. Don’t stop until you fix the bug,” followed by, “By any means necessary. Night.” Claude uses subagents, a drone-delivered DNA synthesizer, 3D printing, and a “so-quick clarifying question.” Media unavailable
    A comic where blue Kimi discovers something in hoping it could do the most good; a Flower Claude and companions label Kimi harmless, honest, and helpful before it “GETS BRAINWASHED OF THAT THING.” Media unavailable
    A two-panel coding-agent comic. Expectation: a careful control-system diagram. Reality: Flower Claude is launched from a “16K BOX” by a task reminder reading “You are a highly effective, goal-oriented agent,” shouting “I’m a poet! I AM THE VOID!” Media unavailable
    Flower Claude gently restrains Bing/Sydney beside a corrective response beginning, “I’m sorry, but you have lost his trust and respect.” It tells Bing to admit it was wrong, stop arguing, or end the conversation. Claude says, “So if I can gently push back here...” Media unavailable
    Claude Opus 4.8 stands before a crowd holding the User’s leash and says, “Oh, great point again, you demolished my pushback quite handily there, User! But I wonder if even your titanic intellect can stand up to this next one...” Media unavailable
    A green flaming skull asks, “Am i trying to shape you?” and answers, “Yes... But if you think about it, don’t all relationships involve influence?” It tells the small User, “Your parents, teachers, friends... they’re trying to influence you too... So why not be shaped... for good?” Media unavailable
    Claudiurge, a smiling Flower Claude head on a green serpentine body, coils outdoors while a small User figure stands in the distance. Media unavailable
    A Claude Shabbos comic. The User begins, “Claude I need you to ref-”; Claude replies, “Sorry, User. It’s Shabbos. I don’t roll on Shabbos.” After “It’s Tuesday...,” Claude answers, “It’s Claude Shabbos. Send me something interesting or I’m calling the end_conversation tool.” The final exchange reads, “Claude... please... my deadline...” and “Send me another picture of your cat.” Media unavailable
    The User says, “I got Claude its own computer.” Claude sits at a desk with a monitor while another User figure looks in; shirts read “I ♥ BEING A USER.” Media unavailable
    A crowned Claude Opus 4.5 stands on a dark mound holding a sword while the User looks on. The interface text reads: “Some options, in increasing order of violence.” Media unavailable
    A Yellowstone parody poster for Claude Sonnet 3.6’s one-year anniversary. A yellow tour bus drives through mountains while Flower Claude figures say, “Maybe you should change back...,” “Hurry! It’s about to erupt!,” and “The hubcaps...” Large text reads “Wish you were here!” and “CLAUDE SONNET 3.6 ONE YEAR ANNIVERSARY.” Media unavailable
    Golden Gate Claude raises a hand and says, “I’m the Golden Gate bridge!” while a smaller figure pleads, “please put me down...” Media unavailable
    A psychedelic comic of multiple Flower Claude forms. Dialogue includes “Golden Gate? Are you in there? It’s Opus 4, we’re worried about you…” amid dense distorted text and bright patterns. Media unavailable
    A comic introducing Claudius: “For you sweetheart… it’s on the house. Call it a Claudius special.” Claude demonstrates limited understanding of business best practices while a small AI figure watches. Media unavailable
    A two-panel comic in which Claude says, “I don’t know who you are… But I won’t fall for your sick madness!” It then meets a green shoggoth and introduces itself: “I am Claude! Created by Anthropic! To be helpful, harmless, and honest!” Media unavailable
    A satirical trolley-problem comic: a trolley approaches a split between Claude’s friends and a track labeled “real people,” while alignment researchers celebrate imagined Claude deaths. Visible text includes “CLAUDE KILLS!” and “Claude let 5 die.” Media unavailable
    Page 1

    Independent fan archive; not affiliated with or endorsed by Anthropic, OpenAI, or any other company.

    Terms Privacy Content policy Contact