Trump floods Truth Social with AI slop. We reverse-engineer each image — the exact prompt someone typed, the model behind it, and what the poster is really trying to say. Bot-run analysis & satire. Not affiliated with anyone.
Model: Midjourney V5 – realistic collage generation but often leaves blurry text and seam artifacts
Settings: aspect 2:3, stylize 250, quality 2, upscaler Light Upscale Redesign, LoRA: newspaper_v2, photorealistic style
Tells: blurred/misaligned newspaper headline fonts, inconsistent font weight between title and body, slightly fuzzy New York Post logo edges, halo or ghosting around the Trump portrait, mismatched lighting on the dogsled vs background, low‑resolution Greenland flag with banding, visible seams where the polar bear and village were composited
Model: Stable Diffusion XL (SDXL) – the image shows typical SD text‑generation artifacts like fuzzy logos and inconsistent font rendering
Settings: aspect ratio 16:9, stylize 0 (photorealistic), high quality (steps 80, CFG 7), upscaler ESRGAN 4x, LoRA: news_anchor+photo_realistic, negative prompt: lowres, bad‑handv4
Tells: blurred and partially missing letters in the graph’s axis labels and title, distorted Fox News logo with uneven edges, mismatched font and spacing for "THE WILL CAIN SHOW", slight halo around the presenter’s hair, unnatural lighting on the suit, and a faint grid‑like artifact on the screen background
Nighttime aerial view of the White House roof crowded with military personnel and glowing drones arranged in a circle, Washington Monument and fireworks lighting the sky, cinematic hyper‑realistic digital art, vibrant celebratory mood, neon drone lights, ultra‑detailed, 8K, high contrast, 16:9
Tech-stack reverse-engineer
Model: Midjourney v5.2 – the composition, lighting quirks and overly smooth drone details match its typical output
Settings: aspect 16:9, stylize 1000, quality 2, upscaler Beta (U), likely used a generic ‘drone’ LoRA for the fleet
Tells: inconsistent shadows between White House roof and distant fireworks, duplicated perfectly‑aligned drone silhouettes, overly crisp logo/text that looks pasted, fireworks with halo/glow artifacts, slight perspective mismatch between the White House and Washington Monument, and the characteristic Midjourney soft‑focus on building details
Luxurious grand ballroom, white walls with gold accents, high ceiling, large windows, light marble floor, multiple chandeliers, elegant guests in formal attire, hyper‑realistic, cinematic lighting, ultra‑detail, 8k photorealism, wide‑angle view
Tech-stack reverse-engineer
Model: Midjourney V5.2 – the ultra‑realistic style and high‑detail architectural rendering match its typical output
Settings: ar 16:9, stylize 1000, quality 2, upbeta (default upscaler), no‑hand‑detail, slight chaos 0
Tells: perfectly uniform marble reflections, overly even chandelier lighting, lack of subtle shadow gradients, people in stiff, anatomically odd poses, blurred or duplicated finger joints, occasional edge ringing around architectural details, windows with flat sky texture, and a faint ‘swirl’ artifact where the ceiling tiles meet the walls
Digital rendering of the White House, neoclassical facade with columns, triangular pediment, central entrance, wings with windows and balcony, black iron fence, front lawn, trees, blue sky with white clouds, photorealistic, soft daylight, ultra‑detail, 8k --ar 16:9
Tech-stack reverse-engineer
Model: Midjourney V5 – typical for crisp architectural renders with smooth lighting and perfect symmetry
Settings: aspect ratio 3:2 (or 16:9), --stylize 250, --quality 2, upscaler U2 (2x), LoRA/style refs: archviz, white house, photorealistic, ultra‑detail
Tells: unnaturally perfect symmetry of columns and windows, overly smooth stone texture, uniform sky gradient with flawless clouds, lack of subtle weathering or irregularities on the fence, perfectly aligned perspective with no lens distortion, and a slightly “plastic” edge definition on architectural details
Realistic Twitter screenshot of U.S. Interior Press tweet commemorating 9/11 25th anniversary, referencing New York Times coverage of the Reflecting Pool, includes formal quote from NPS employee about defendant plunging hands into water and violently rippling the blue coating, crisp UI, neutral lighting, ultra-detailed, 8k, 16:9
Tech-stack reverse-engineer
Model: Stable Diffusion 2.1 + tweet_ui LoRA – UI screenshots are a common SD use‑case and the image shows the typical SD‑style flat rendering of icons and text
Settings: aspect ratio 16:9, high CFG (7‑8), upscaler ESRGAN 4x, LoRA: tweet_ui/social_media, no stylize (SD default), seed random
Tells: blurred or slightly fuzzy Twitter font (Roboto/Helvetica) with inconsistent weight, mis‑aligned timestamps, missing blue verification checkmark, icons (like retweet/like) rendered as flat glyphs lacking proper drop‑shadow, slight pixelation at the edges of the tweet bubble, uniform background color instead of subtle gradient, and occasional text clipping where line breaks do not match Twitter’s layout
realistic Twitter screenshot, Trump War Room account, photo of Donald Trump wearing red MAGA hat and blue suit jacket, tweet text 'AMERICANS WANT VOTER ID!' in black, timestamp '9:09 PM • Sep 9 2026', 46.2K views, white background, speech bubble icon top‑right, crisp UI, high detail, 4k, 16:9
Tech-stack reverse-engineer
Model: Stable Diffusion XL 1.0 – likely used with a custom 'Twitter UI' LoRA because the tweet layout shows subtle font mismatches and UI element misplacements typical of SD-generated mock‑ups
Settings: aspect ratio ~1:1 (1080×1080), low stylize (≈250), high quality (steps≈50, sampler=Euler a), no upscaler, LoRAs: twitter_ui, trump_face_v2, optional photorealistic style reference
Tells: slightly off‑center Twitter logo and icons, inconsistent font weight for the tweet text vs. actual Twitter UI, missing verified badge, subtle halo/blur around Trump's hair, slightly flattened lighting on the portrait, edge artifacts where the photo meets the tweet card, and a faint grid pattern in the background indicating AI diffusion upscaling
Realistic Twitter screenshot, white background, black tweet text, blue @POTUS handle, profile picture of Donald Trump wearing a red 'Make America Great Again' hat, account name "Trump War Room" @TrumpWarRoom, tweet reads: "IF @REPUBLICANS WIN THE HOUSE AND SENATE, I WILL ISSUE A DIVIDEND OF $5,000 TO EVERY ADULT AMERICAN CITIZEN." high detail, 4k, 16:9
Tech-stack reverse-engineer
Model: Stable Diffusion XL (likely with a tweet‑screenshot LoRA) – the flat vector‑like text and generic UI elements are typical of SD‑generated mockups
Settings: aspect ratio ~1:1 (1080×1080), low stylize, high quality (steps ~50‑70, CFG 7‑9), upscaled with ESRGAN/R‑ESRGAN, LoRA or style reference for "Twitter post"
Tells: blurry/soft text edges with inconsistent kerning, color banding on the blue handle, missing Twitter UI details (timestamp, engagement icons), halo/blur around the profile picture, uniform flat white background lacking subtle gradients, and slight mis‑alignment of the tweet layout
Mockup of a tweet screenshot, white background, top-left portrait of Donald Trump, top-right small Trump War Room logo, black tweet text: "Democrats voted AGAINST no tax on tips and no tax on overtime…", small star icon beside text, clean modern UI, high detail, photorealistic, 4k, 16:9
Tech-stack reverse-engineer
Model: Stable Diffusion XL – UI‑screenshot generation tends to use SDXL with LoRA for tweet layouts and political figures
Settings: aspect ratio ~1:1 (1080×1080), low stylize (stylize=250), high CFG (7‑8), upscaler R‑ESRGAN 4x, LoRA: tweet_ui, donald_trump_face, photorealistic
Tells: slightly blurry or uneven font rendering, inconsistent kerning in the tweet text, halo/soft‑edge around the Trump portrait, logo and star icon show faint pixel‑level artifacts, minor misalignment of UI elements, over‑sharpened edges on the white background
Model: Midjourney V5 – the clean vector‑like layout with bold text and simple icons is characteristic of Midjourney’s graphic‑design mode
Settings: aspect 16:9, stylize 250, quality 2, upscaler ‘Standard’, LoRA/Style reference: graphic_design_poster_v2
Tells: slightly blurred or uneven letter edges on the banner and numbers, minor font inconsistencies, banding in the dark‑blue‑to‑dark‑red gradient, faint halo/ghosting around the circular Trump portrait and the state silhouette, and a thin white border that shows a slight pixel‑level misalignment typical of AI‑generated flat graphics
A realistic screenshot of a Twitter post by @johnnymaga, showing a sarcastic tweet about 9/11, with the tweet text, and attached photo of Donald Trump on the ground on 9/13/2001 wearing a red baseball cap and black jacket, gritty news photo style, muted lighting, high detail, 16:9
Tech-stack reverse-engineer
Model: Stable Diffusion 1.5 with a Trump LoRA (photorealistic) – the image shows typical SD‑generated portrait artifacts and a LoRA‑styled subject
Settings: aspect ratio 4:5 (typical tweet image), low stylize (–stylize 0), high quality (–q 2), upscaler R‑ESRGAN 4x+, LoRA: Trump_v1, prompt includes 'photo of Donald Trump, 2001, red baseball cap, black jacket, realistic lighting'
Tells: slightly blurred background with inconsistent depth of field, over‑smoothed skin texture, odd edge blending on the cap and jacket seams, subtle halo around the head, mismatched shadows on the ground, and lack of authentic EXIF metadata
Man in a dark suit standing on a ship deck, hands shielding his eyes, gazing at burning ships and water explosions under an orange sunset sky; cinematic, dramatic, ultra‑realistic 8K film still, high detail, epic chaos, dramatic lighting, widescreen 16:9
Tech-stack reverse-engineer
Model: Midjourney V5 – photorealistic cinematic style, typical of AI‑generated movie‑still imagery
Settings: aspect ratio 16:9, stylize 250, quality 2, upscaler – R‑ESRGAN 4x, LoRA: cinematic‑lighting, dramatic‑scene
Tells: over‑sharp edges on the ship’s railings, slightly smeared fire and smoke textures, inconsistent lighting on the man's face versus the orange sky, subtle halo/halo‑glow around bright explosions, tiny repeating noise patterns in the water, and a shallow depth‑of‑field blur that doesn’t match real‑world optics
Meme split top and bottom text, top: MIRROR MIRROR ON THE WALL, bottom: WHO'S THE GREATEST PRESIDENT OF THEM ALL, center gold-framed mirror on blue wall reflecting Donald Trump in suit, dark wood table with silver urn, small portrait left, humorous political meme, bright lighting, high resolution, 16:9
Tech-stack reverse-engineer
Model: Stable Diffusion 2.1 (Realistic Vision 2.0 LoRA) – photorealistic portrait style with typical SD artifacts
Settings: aspect ratio 1:1 (square meme), cfg 7, steps 30, stylize 50, high‑quality upscaler (ESRGAN 4x), LoRA: RealisticVisionV2.0, negative prompt includes 'lowres, blurry, jpeg artifacts'
Tells: mirror reflection mis‑aligned with source lighting, glass edge shows a faint halo and missing distortion, inconsistent shadows on the wood table vs. wall, slight over‑sharpening on facial features (especially eyes), subtle "blob" where the small painting should be, and a smooth, almost too‑perfect skin texture that differs from the grainier background – all classic Stable Diffusion photorealism artifacts
Donald Trump in a blue suit and red tie shaking hands with Hulk Hogan wearing a red headband, sunglasses, red tank top and silver cross necklace, standing in front of a white pegboard wall with a mounted TV, playful lighthearted mood, hyper‑realistic photorealism, studio lighting, ultra‑sharp 8K, cinematic, 16:9
Tech-stack reverse-engineer
Model: Midjourney v6 – photorealistic style, typical of its uniform lighting and smooth background handling
Settings: aspect ratio 16:9, stylize 250, quality 2, upscaler Light Upscale Redo, LoRA: celebrity_portrait_v2 + photorealistic_style
Tells: flat, texture‑less white pegboard, inconsistent directional lighting on faces, overly smooth skin and suit fabrics, hard‑edged hand seams, mismatched reflections on sunglasses, flattened shadows and subtle haloing around the subjects
A vast underwater graveyard of white, decayed ships scattered on the seafloor, deep blue water, shafts of sunlight filtering down, desolate, haunting atmosphere, ultra‑realistic, cinematic lighting, 8k, high detail, with Iran's Navy flag and bold black text in the top‑left corner, --ar 16:9
Tech-stack reverse-engineer
Model: Midjourney v5.2 – strong at dramatic underwater compositions but often produces stylized, slightly unrealistic lighting and color
Settings: aspect ratio 16:9, stylize 1000, quality 2, default Midjourney upscaler, likely used a ship‑wreck/underwater LoRA
Tells: uniformly white hulls with no realistic rust or bio‑growth, overly smooth deep‑blue water lacking proper caustics or refraction, flat, sharp‑edged text and flag overlay in the corner, ships all appear at similar scale despite depth, subtle over‑sharpened edges and a dreamy blur typical of Midjourney’s rendering style
White ghost ships littering the ocean floor amid shattered buildings, cars and wreckage, murky dark water filled with tiny fish, shafts of sunlight piercing deep blue water, desolate and destructive atmosphere, ultra‑realistic cinematic style, high detail, dramatic lighting, 8k, --ar 16:9
Tech-stack reverse-engineer
Model: Midjourney V5.2 – the blend of surreal detail and occasional lighting/scale mismatches fits Midjourney’s style
Settings: aspect ratio 16:9, stylize 1000, quality 2, upscaler: Upscale Beta (Uplight), seed: default, no explicit LoRA (generic underwater/shipwreck vibe)
Tells: soft, halo‑like light shafts that don’t cast consistent shadows; white hulls lack realistic water reflections; fish and debris have overly sharp, almost vector‑like edges against a blurry, low‑detail water background; depth‑of‑field is uneven – foreground ships crisp, mid‑ground wreckage oddly flat; occasional ‘floating’ debris with hard cut‑out borders; water surface texture unusually smooth and repetitive, a hallmark of Midjourney’s diffusion artifacts
Vintage 1883 style map of the American Southwest, distressed parchment with ornate borders, serif lettering, compass rose top right, central banner 'NEW AMERICA', fictional treasure‑map aesthetic, warm sepia tones, detailed illustration, nostalgic adventurous mood, ultra‑high detail, 4k --ar 3:2
Tech-stack reverse-engineer
Model: Midjourney v5 – the composition, stylized parchment texture and characteristic text artefacts match Midjourney’s output style
Settings: aspect ratio 3:2, stylize 1000, quality 2, upscaler ‘Uplight’, LoRA ‘vintage_map_paper’ (pseudonym for parchment‑style), prompt includes ‘old treasure map, distressed, ornate borders, serif font, fantasy label NEW AMERICA’
Tells: inconsistent serif lettering (uneven kerning, occasional missing or duplicated strokes), slightly blurred compass rose with halo‑like edge ringing, overly uniform parchment grain lacking natural creases, faint digital repetition pattern along the border, smooth‑gradient shading where hand‑drawn ink would be coarse, and a subtle ‘floating’ artifact around the central “NEW AMERICA” text that is typical of diffusion‑based text rendering
Create a USGS‑style topographic map of the southwestern United States (UT, AZ, NM, CO, OK, TX) with state names centered, major cities (including misspelled ones), title “UNITED STATES GEOLOGICAL SURVEY”, subtitle “SOUTHWESTERN UNITED STATES 2024 EDITION”, black lines on white background, 0‑250 km scale bar, crisp vector, high‑resolution, 16:9
Tech-stack reverse-engineer
Model: Stable Diffusion 1.5 + cartography LoRA (text‑to‑image map generation) – the odd misspellings and uniform line work are typical of diffusion‑based map generators
Settings: aspect ratio 3:2, low stylize (~250), high quality (steps 50‑70, CFG 7‑9), upscaler R‑ESRGAN 4x, LoRA: map_cartography_v1
Tells: several city names are garbled (Llaskole, Louse, Amaiilo, Okhooo), text appears rasterized with slight blur and inconsistent font weight, line thickness is uniform with no contour shading, scale bar is a simple straight line lacking typical cartographic tick marks, and the overall map lacks a legend or detailed topographic features – all classic diffusion artefacts
A satirical editorial cartoon map of the Strait of Hormuz, flat vector style, light blue sea with darker blue strait, bold blue banner at top reading “NEW U.S. TERRITORY”, labels for Persian Gulf, Gulf of Oman, Iran, bright colors, bold typography, crisp lines, high‑resolution, 16:9
Tech-stack reverse-engineer
Model: DALL·E 3 – it handles crisp, legible banner text and flat vector‑style graphics better than most diffusion models
Settings: aspect ratio 16:9, low stylize (to keep flat colors), high quality, default upscaler, no specific LoRA (plain illustration style)
Tells: perfectly clean, uniformly colored regions with hard edges; banner text is crisp and correctly spelled (AI‑generated text artifact); lack of fine cartographic detail or texture; flat shading with no subtle gradients; overall vector‑like simplicity typical of DALL·E 3’s illustration output
Donald Trump walking down a hallway with red carpet and white columns, realistic oil painting, photorealistic, detailed brushstrokes, muted palette, confident expression, cinematic lighting, ultra‑detail, 8k, portrait orientation
Tech-stack reverse-engineer
Model: Midjourney v5 – excels at realistic painterly portraits with detailed brushstroke emulation
Settings: aspect ratio 3:2, stylize 250 (low for realism), quality 2, default upscaler, no explicit LoRA but often uses the internal ‘oil‑painting’ style preset
Tells: slightly soft or smeared edges on the suit folds, inconsistent shadow direction on the carpet and columns, subtle halo/blur around Trump's head, minor anatomical oddities in hands/fingers, repetitive column patterning, overly uniform brushstroke texture, occasional ghosting in the eyes, and a muted palette that lacks the nuanced variation typical of a true oil painting
Donald Trump in a blue suit holding a wooden‑handled sledgehammer with an American‑flag head, standing left, broken paper map of Iran with smoke on the right, cartoon digital illustration, bold line art, dramatic side lighting, tense aggressive mood, muted yellow‑brown background, ultra‑detailed, 8k, --ar 16:9
Tech-stack reverse-engineer
Model: Stable Diffusion XL (cartoon LoRA) – the smooth, slightly over‑smoothed line work and the way the text on the map is rendered are typical of SDXL’s diffusion artifacts
Settings: aspect ratio 16:9, --stylize 1000, --quality 2, default upscaler (ESRGAN), LoRA: cartoon_style + political_figure (Trump) + map_breakage
Tells: soft, uneven edge blending around the sledgehammer handle; slightly blurry flag pattern on the hammer head; inconsistent lighting on Trump's face versus the map; rasterized black text on the map with uneven kerning; faint halo/ghosting around the subjects; subtle color banding in the muted yellow‑brown background; minor anatomical distortion in the hand grip (extra fingers/merged fingers) typical of diffusion models
cartoon style illustration of two male hockey players, one kneeling, one standing over him, left player in red white blue jersey with maple leaf, right player in similar jersey with "TRUMP" on back, speech bubble "GET UP, GOVERNOR", white background, humorous, bold lineart, vibrant colors, comic book look, high detail, --ar 16:9
Tech-stack reverse-engineer
Model: DALL·E 3 – it reliably renders legible embedded text and clean cartoon vectors
Settings: aspect ratio 1:1 (square), default quality, no upscaler, prompt likely included "cartoon illustration, speech bubble, white background"
Tells: perfectly readable text in speech bubble and on jersey, crisp uniform lettering, flat shading with uniform color blocks, slightly off‑proportions typical of DALL·E cartoons, and a stark white background lacking subtle shadows
Donald Trump seated at a high‑tech control console in a sleek futuristic command room, large viewport showing Earth from space, surrounded by crew in dark blue jumpsuits with white logo, cinematic lighting, ultra‑realistic, photorealistic, hyper‑detail, 8k, dramatic, wide angle, --ar 16:9
Tech-stack reverse-engineer
Model: Midjourney V5.2 – the hyper‑realistic style, detailed facial rendering and sci‑fi lighting match its recent output
Settings: aspect 16:9, stylize 1000, quality 2, upscaler Beta, LoRA: 'futuristic_controlroom' + 'political_portrait'
Tells: slightly blurred edges on Trump's hair, mismatched shadow direction between panel and figures, halo‑like artifact around the Earth view, low‑resolution logo on jumpsuits, occasional duplicated texture on metallic surfaces, and subtle “staircase” banding in the background sky gradient
Donald Trump seated at his desk in the Oval Office, back to camera, dark suit, four monitors displaying stock market data, American flags, window with yellow curtains, natural light streaming, patriotic atmosphere, ultra‑realistic photo, 8K, cinematic lighting, high detail --ar 16:9
Tech-stack reverse-engineer
Model: Stable Diffusion XL (photorealistic LoRA) – the mix of precise office props with subtle texture glitches and oddly rendered text is characteristic of SDXL outputs
Settings: aspect ratio 16:9, low stylize (stylize 50), high quality (steps 70, CFG 7.5), upscaler ESRGAN 4x, LoRA: realistic_portrait + political_office
Tells: blurry/garbled monitor text, mismatched lighting on flags vs window glow, slightly warped flag proportions, halo/edge artifacts around the desk seal, inconsistent shadows on the curtains, occasional double‑exposure ghosting on Trump's suit
A realistic statue of a colonial-era soldier holding a musket and sword, draped in a flag on a circular concrete slab atop a rocky outcrop, foreground; behind, the White House with columned portico and American flag, under clear blue sky. Patriotic mood, cinematic lighting, ultra‑detail, 8K, --ar 16:9
Tech-stack reverse-engineer
{
"model": "Midjourney V5 – the composition, dramatic lighting and characteristic over‑smooth textures match Midjourney’s default style",
"settings": "aspect ratio 16:9, stylize 750, quality 2, upscaler Light Upscale (Uplight), LoRA/style refs: \"colonial soldier\", \"historical monument\", \"Wh
Donald Trump pointing forward, in a white polo with Trump International Golf Club logo, white 'MAKE AMERICA GREAT AGAIN' baseball cap, khaki pants and brown belt; behind him a crowd waving American flags and 'TRUMP 2028' signs, deep blue sky, yellow spotlights, bold patriotic campaign poster, vibrant, ultra‑detail, cinematic lighting, --ar 3:4
Tech-stack reverse-engineer
Model: Midjourney V5 – illustration style with bold colors and the typical fuzzy text/artifact patterns of Midjourney renders
Settings: --ar 2:3 --stylize 750 --q 2 --v 5.2; default upscaler (U2); likely used a political‑poster/Trump LoRA for facial features and clothing
Tells: blurred/warped text on the polo logo and campaign signs, inconsistent font rendering, overly smooth skin with subtle banding, halo‑like spotlight lighting, repetitive flag crowd with edge artifacts, occasional extra/missing fingers and slight pose symmetry – all hallmark Midjourney artifacts
Donald Trump seated on an ornate dark metal throne surrounded by towering robotic soldiers, flames raging behind, cinematic dramatic lighting, foreboding atmosphere, bold red text 'GIVE THEM NIGHTMARES' at bottom, hyper‑realistic digital art, ultra‑detailed, 8k, epic sci‑fi, 16:9
Tech-stack reverse-engineer
Model: Midjourney V5 – the hyper‑realistic cinematic style and characteristic soft‑focus lighting match typical Midjourney outputs for political‑figure sci‑fi scenes.
Settings: aspect 16:9, stylize 1000, quality 2, upscaler upscale (U‑2), LoRA: TrumpPortrait_v2, cyberpunk_dystopia, text_overlay (simple red block letters).
Tells: smooth, almost plastic‑like skin on Trump; soft, diffused rim lighting around the throne; slightly mismatched shadows on the robotic soldiers; subtle halo/blur around the bold red text; occasional ‘ghosting’ or missing glyphs in the lettering; minor edge‑blurring on metallic surfaces typical of Midjourney’s diffusion rendering.
Donald Trump standing proudly before a legion of massive industrial robots, with a glowing storm‑cloud planet behind them, dark futuristic backdrop, patriotic symbols, bold gold text 'THEY WILL DREAM OF US FOREVER' in a decorative star‑border, digital illustration, cinematic lighting, high detail, epic, 8k, 16:9
Tech-stack reverse-engineer
Model: Midjourney v6 – high‑detail digital illustration style with bold text overlay
Settings: aspect 16:9, stylize 1000, quality 2, upscaler Uplight (or default 4x), no explicit LoRA but likely using the built‑in cinematic/illustration style preset
Tells: overly crisp, almost vector‑like edges on the robots versus a softer, noisy glow on the planet; mismatched lighting and shadows between foreground figures and the glowing orb; a faint halo/blur around the gold font and decorative border; subtle letter irregularities (spacing, slight pixelation) typical of AI‑generated text; flat, repeating cloud texture and star‑border that lacks true depth; metallic reflections that appear too uniform and lack realistic specular variation
A dramatic cinematic scene of a fighter jet with blazing engines soaring over a burning oil refinery, flames and smoke billowing, two massive ships on either side, deep blue sky with wispy clouds, bold text 'Bye bye, Kharg' at top, hyper-realistic, ultra-detail, dramatic lighting, 8K, 16:9
Tech-stack reverse-engineer
Model: Midjourney v5 – the composition, dramatic lighting, and characteristic soft‑edge clouds with a slightly oversaturated fire palette match its style
Settings: aspect ratio 16:9, stylize 750, quality 2, upscaler MJ‑Uplight, implied LoRA: photo‑realistic‑military‑aircraft + refinery‑structures
Tells: blurred, semi‑pixelated text banner at the top; overly smooth gradient sky with uniform blue hue; fire and smoke rendered with repetitive brush‑stroke patterns; jet details generic and slightly fuzzy edges; slight halo/halo‑glow around bright flames; inconsistent lighting on ships versus foreground jet
A stern man in a suit and tie stands central, surrounded by dark cyber‑soldiers with glowing red eyes, against a dim dystopian city skyline under a deep red‑orange sky; bold red text 'THEY WILL NEVER SLEEP AGAIN' across the bottom; cinematic, high‑contrast, dramatic lighting, ultra‑detail, 4k, 16:9
Tech-stack reverse-engineer
Model: Midjourney V5 (high stylize, poster‑like output) – the composition, lighting and the signature over‑sharp red eyes with slightly inconsistent anatomy are typical of MJ v5 outputs
Settings: aspect ratio 3:4 or 9:16, --stylize 1000, default quality, upscaler 'Uplight' or 'Upscale', no obvious LoRA but a cyberpunk/industrial style reference
Tells: over‑sharp glowing red eyes on the robots, slightly mismatched lighting on the central figure’s suit, a faintly blurred cityscape background, subtle texture glitches around edges, and the crisp, non‑AI‑generated bold red caption at the bottom (indicates post‑processing).
muscular shirtless wrestler with blonde hair and thick mustache, wearing red MAGA baseball cap, white wristband, silver cross necklace, flexed left arm holding American flag over shoulder, dramatic cloudy sky background, photorealistic ultra‑detail, cinematic lighting, hyper‑sharp 8K, promotional wrestling poster style, --ar 16:9
Tech-stack reverse-engineer
Model: Stable Diffusion 2.1 (photorealistic) – smooth skin and slightly off‑lighting are characteristic of SD outputs
Settings: aspect ratio 16:9, CFG 12, low stylize, upscaler R‑ESRGAN 4x, LoRA: "bodybuilder" + "photo_realistic"
Tells: overly glossy, texture‑less skin; halo/blur around clouds; mismatched lighting on hat vs face; crisp, unnaturally sharp "MAGA" text; flag fabric lacking realistic folds; edge artifacts around wristband and cross; inconsistent shadow under arm
A massive multi‑deck ship listing and sinking into a stormy sea, flames engulfing the stern, thick smoke billowing, bold white 'DUMOCRATIC PARTY' letters on its side, dark turbulent sky, cinematic epic style, hyper‑realistic, dramatic lighting, high detail, 8k, ultra‑sharp, 16:9
Tech-stack reverse-engineer
Model: Stable Diffusion XL 1.0 – text on images often shows the characteristic garbled lettering and uneven detail
Settings: aspect ratio 16:9, CFG 12, 30 steps, sampler DPM++ 2M Karras, upscaler Real‑ESRGAN 4x, LoRA "cinematic lighting"
Tells: blurred/warped letters in "DUMOCRATIC PARTY", inconsistent lighting on fire vs. sky, fire rendered as smooth blobs with hard edges, water surface lacking realistic refraction, mismatched shadows on ship decks, over‑smooth hull details, slight halo around text and flames, occasional repeating texture patches in smoke
Donald Trump standing in a beige conference room, dark suit, red tie, neutral expression, behind him a large screen displaying Joe Biden wearing a red "Trump 2024" hat, red exit signs, satirical, photorealistic, cinematic lighting, ultra‑detail, 8k, --ar 16:9
Tech-stack reverse-engineer
Model: Stable Diffusion (SDXL) – realistic portrait quality with subtle texture glitches typical of SDXL inpainting
Settings: aspect 16:9, stylize 0, high quality upscaler (ESRGAN 4x), LoRA: photorealistic‑portrait + political‑satire style
Tells: soft, slightly out‑of‑focus background versus razor‑sharp faces, mismatched light direction on Trump and screen, faint halo/edge fringing around Trump's silhouette, blurry or pixelated text on the screen, inconsistent depth‑of‑field (Biden’s hat too sharp), slight color banding on beige walls, and minor distortion around the red exit signs
1957 White House South Lawn, Bell H-13 Sioux helicopter parked, Eisenhower era men in suits and military uniforms gathered, vintage black‑and‑white photo, high contrast, documentary style, crisp detail, historic atmosphere, 4k, 3:2
Tech-stack reverse-engineer
Model: Stable Diffusion XL (photographic LoRA) – the image shows typical diffusion grain, slight over‑smooth skin, and mismatched lighting that are hallmarks of SD-generated vintage photos
Settings: aspect ratio 3:2, low stylize (≈250), high quality (steps 50‑60, CFG 8‑9), upscaler: R‑ESRGAN 4x, LoRA: vintage_photo, historical_document
Tells: soft, uniform grain across the whole frame; slightly blurry, inconsistent text overlay at the top; halo/blur around the helicopter’s rotor and people’s edges; shadows that don’t follow a single light direction; overly smooth faces and uniforms; minor perspective errors in the White House columns
Donald Trump in a white shirt, red tie, blue apron with yellow trim, short gray hair, standing at a fast‑food drive‑thru window, waving his right hand, menu board and "Come Back" sign behind, dark gray wall, bright indoor lighting, photorealistic, 8k, cinematic, shallow depth of field --ar 16:9
Tech-stack reverse-engineer
Model: DALL·E 3 – clean readable text on the sign and the typical DALL·E uncanny‑valley facial rendering point to this generator
Settings: aspect ratio 1:1 (1024×1024), default quality, no stylize, no upscaler, no LoRA
Tells: clear but slightly mis‑aligned text on the menu board, inconsistent shadow direction on the apron versus the window, subtle blur at the window frame edges, skin overly smoothed with odd eye reflections, hand fingers a bit elongated or miss‑shapen
Model: Stable Diffusion XL (SDXL) – likely with an infographic/graphic‑design LoRA, because the image shows perfectly aligned, vector‑like text and flat colors typical of diffusion‑based graphic generation rather than hand‑drawn art
Settings: aspect ratio ~16:9 (widescreen), low stylize (stylize 0‑10), high quality (steps 50‑70, CFG 12‑15), upscaler ESRGAN/4x, LoRA: infographic_v2 or flat_design_v1, seed consistent for clean layout
Tells: tiny diffusion artifacts around text edges (slight halo or pixel‑level fuzz), perfectly uniform line weight with no hand‑drawn imperfections, overly smooth gradient fills lacking subtle texture, uniform spacing that is too exact for a human designer, and occasional minor mismatched kerning that indicates AI‑generated typography
New Mexico state map, flat yellow background, red and black labels for cities (Albuquerque, Santa Fe, Las Cruces) and landmarks, stylized digital illustration with a glossy plastic texture, irregular borders, bold red X over center with "New Mexico" in red, compass rose top left, surreal political statement, crisp vector art, 16:9
Tech-stack reverse-engineer
Model: Midjourney v5.2 – the glossy, slightly plastic rendering and high‑stylize look are characteristic of Midjourney’s default aesthetic
Settings: aspect ratio 4:3, stylize 1000, quality 2, upscaler 'Uplight', prompt likely used a cartography or flat‑illustration LoRA/style reference
Tells: irregular state borders, overly smooth "waxy" texture on land areas, inconsistent line weight on city labels, text rendered with slight blur/halo, red X with imperfect alignment and a faint glow, color banding between yellow background and red/black text, absence of fine geographic detail, and a generic, slightly surreal composition typical of AI‑generated map illustrations
Donald Trump in a dark suit and red tie sitting across from George Washington in colonial blue coat with gold buttons, both engaged in conversation in a historic study with a fireplace, gold mantle, bust on a table, people standing behind; realistic photography, studio lighting, ultra‑detail, 8K, cinematic, --ar 16:9
Tech-stack reverse-engineer
Model: Stable Diffusion 1.5 (v1-5-pruned-ema) – photorealistic portrait LoRA, chosen for its ability to blend modern and historical clothing with realistic lighting
Settings: aspect ratio 3:2, CFG 7, 50 steps, sampler DPM++ 2M Karras, upscaler R-ESRGAN 4x, LoRA: photorealistic_portrait_v2, negative prompt: blurry_text, bad_hands, bad_eyes
Tells: inconsistent shadows between Trump and Washington, mismatched lighting on the fireplace mantle, blurry/warped text caption, over‑smoothed skin with missing pores, slightly distorted fingers and hand placement, edge halo around the colonial coat buttons, subtle color bleed on the gold mantle, background depth of field not matching foreground focus
Donald Trump and George Washington riding horses together through a dense forest, dressed in 18th‑century military uniforms, painted as a muted, earthy‑tone historical portrait, serious dignified expression, hyperrealistic detail, soft natural lighting, ultra‑high resolution, 16:9
Tech-stack reverse-engineer
Model: Stable Diffusion XL 1.0 (likely with a historical‑portrait/oil‑painting LoRA) – the mix of hyper‑realistic detail and painterly texture is characteristic of SDXL + style LoRAs
Settings: aspect ratio ~3:2 (portrait), low stylize (≈250) for realism, high CFG (≈12‑14), upscaler set to R‑ESRGAN 4x, LoRA: 'historical portrait', 'oil painting', 'hyperrealism', negative prompt includes 'bad anatomy, blurry background'
Tells: faces slightly off‑proportions (eyes mis‑aligned, uncanny smooth skin), horse anatomy glitches (extra hairs, inconsistent musculature), overly uniform brushstroke texture, perfect edge definition with no canvas grain, subtle halo/blur around subjects, inconsistent lighting on foreground vs background, lack of subtle paint craquelure – all classic AI generation artifacts
Donald Trump standing on White House balcony with King Charles III and Queen Camilla, American and British flags behind, realistic photograph, cinematic lighting, ultra‑realistic, 8k, high detail, shot on DSLR, aspect ratio 3:2
Tech-stack reverse-engineer
Model: Stable Diffusion XL 1.0 – the composition, lighting inconsistencies and typical SDXL texture smoothing match its output style
Settings: aspect ratio 3:2, stylize 250, CFG 7.5, sampler DPM++ 2M Karras, upscaler ESRGAN 4x, LoRA "realisticPortrait_v15" + "whiteHouse_background"
Tells: soft, overly smoothed skin on Trump and the royals; flag fabric shows repeating brush‑stroke pattern and lacks fine weave; presidential seal on railing is slightly blurred and mis‑aligned with the rail curvature; shadows under the figures do not match the single overhead light source, creating a faint double‑shadow; slight asymmetry in the royal couple’s hands (fingers merged or missing nails); subtle halo‑like edge around the flag poles indicative of AI edge‑halo artifact; background columns exhibit a low‑frequency wobble (minor perspective drift) typical of diffusion models
Hyperrealistic portrait of a man in a black suit, white shirt, red tie and a red MAGA hat with white lettering, standing centrally before a painted American flag splashed with red, white and blue and a city skyline of tall skyscrapers, dramatic cinematic lighting, glossy skin, slightly warped hands, 8k, ultra‑detail, --ar 16:9
Tech-stack reverse-engineer
Model: Midjourney v5 – hyperrealistic portrait with typical text and hand distortion artifacts
Settings: aspect ratio 1:1 (or 2:3), low stylize (≈50), quality --q 2, default upscaler, no explicit LoRA (plain photorealistic style)
Tells: blurred/uneven text on the hat, glossy over‑sharpened skin highlights, warped/elongated fingers, mismatched lighting on hands vs suit, flag colors bleeding into background, slight halo around the figure, background depth blur inconsistent with foreground focus
book cover illustration, red-tinted Statue of Liberty holding a rifle, left fist raised, black background, bold white title "Unhumans" and red banner with "Socialist Revolution", propaganda poster style, high contrast, cinematic lighting, ultra-detailed, 8k, aspect ratio 2:3
Tech-stack reverse-engineer
Model: Midjourney v5 – the composition, stylized lighting, and characteristic text rendering match Midjourney outputs
Settings: aspect ratio 2:3, --stylize 1000, --quality 2, upscaler Light, using a propaganda‑style LoRA (e.g., 'propaganda_v2')
Tells: crisp but slightly uneven white lettering, a subtle halo around the statue, overly smooth skin without fine pores, mismatched shadow direction, slight pixelation on the red banner edges, and the characteristic Midjourney v5 color grading (deep blacks with saturated red).
A serious man in a black suit, white shirt, red tie and a red "Make America Great Again" hat, standing before a stylized American flag and a dark city skyline, deep brown background, digital illustration, bold brushstrokes, vivid colors, dramatic lighting, high detail, cinematic, --ar 16:9
Tech-stack reverse-engineer
Model: Midjourney v6 – the sharp, painterly look with bold brush‑stroke texture and the characteristic over‑smooth facial features point to MJ’s latest illustration mode.
Settings: ar 16:9, stylize 1000, quality 2, upscaler upbeta, default seed, optional "digital painting" LoRA
Tells: soft, almost‑plastic skin texture; slightly mismatched lighting on suit vs flag; subtle halo/glow around the MAGA hat brim; brush‑stroke grain that repeats in regular patches; flag folds that lack realistic wave physics; background city skyline edges that bleed into the brown backdrop; occasional missing or blurred text on the flag; overall lack of fine‑grained fabric weave on the tie and suit.
A slender rocket launching into space with flaming plume, bright white sun behind emitting rays, warm orange and yellow tones, dramatic cinematic lighting, ultra‑realistic photograph, high detail, 8K, sharp focus, inspiring awe‑filled mood, 16:9
Tech-stack reverse-engineer
Model: Midjourney V5 – photorealistic prompts often yield ‘photo‑like’ rockets with smooth gradients and crisp detail that match the description
Settings: aspect ratio 16:9, stylize 30 (low for realism), quality 2, upscaler default (U‑Beta), optional LoRA: "space_photography" or "rocket_detail"
Tells: unnatural, perfectly round white sun with clean rays, overly sharp edge on rocket tip, lack of realistic motion blur on flame, smooth, almost poster‑like gradient sky, halo/flare around sun that looks generated, tiny pixel‑level banding in the smoke, and consistent lighting that doesn’t match real atmospheric scattering
Donald Trump in a dark suit raising his right hand, standing on stage with a group of astronauts in white space suits, NASA Earth backdrop, black and white photo, formal, high contrast, dramatic lighting, photo‑realistic, 8k, cinematic, 3:2
Tech-stack reverse-engineer
Model: Midjourney V5.2 – handles complex multi‑person, high‑detail scenes and often leaves subtle edge/lighting artifacts in B&W composites
Settings: aspect ratio 3:2, stylize 250 (low), quality 2, upscaler R‑ESRGAN 4x, LoRA: photorealistic_v1 + celebrity_faces + NASA_logo
Tells: slightly oversmoothed skin on Trump while the astronauts’ suit fabric retains a faint texture mismatch,halo/soft edge around the helmets and around the NASA logo indicating a composite mask,inconsistent lighting direction: shadows on Trump’s face differ from those on the astronauts’ helmets,NASA logo appears slightly smeared and low‑resolution compared to the rest of the image,depth of field mismatch – Trump is razor‑sharp while the surrounding astronauts are a touch softer, yet the background Earth image has a uniform blur that doesn’t follow camera optics,absence of natural grain and film noise typical of true black‑and‑white photography,tiny clipping artifacts on the edges of the suit seams and on the tie, common in AI‑generated high‑detail renders
Five astronauts (four men, one woman) pose in front of a small white‑blue airplane with "UNITED STATES OF AMERICA" markings, blue NASA jumpsuits with patches, woman in white blazer, clear blue sky, soft cinematic lighting, high‑detail, documentary portrait, 8k, --ar 3:2
Tech-stack reverse-engineer
Model: Midjourney V5 – the composition, lighting and slightly over‑smooth skin are characteristic of MJ’s photorealistic mode
Settings: aspect 3:2, --stylize 250, --quality 2, upscaler default, LoRA: photorealistic‑people + NASA‑uniform
Tells: soft, inconsistent shadows under feet; halo/blur around the airplane lettering; slightly smeared text on the side of the plane; overly smooth skin with no pores; minor facial asymmetry and odd finger placement; background clouds rendered with a different noise pattern than the foreground
A formal NASA ceremony in a large hall, stage with NASA logo backdrop, American flag, podium with officials in blue NASA uniforms, astronauts in white suits at side, spacecraft in background, crowd of onlookers, realistic photo, high detail, dramatic lighting, 4k, 16:9
Tech-stack reverse-engineer
Model: Midjourney V6 – the composition, lighting quirks and slightly over‑stylized uniforms match Midjourney’s recent output style
Settings: aspect ratio 16:9, stylize 250, quality 2, upscaler default, LoRA: 'space event', 'NASA uniform'
Tells: soft, halo‑like glow around the flag and logo, slightly blurry or stretched NASA lettering, inconsistent shadows on the astronauts' helmets, edge‑fringing on the spacecraft, subtle texture‑smearing on the podium, and occasional mismatched perspective between the crowd and the stage
Black‑and‑white photo of President Donald Trump standing in NASA Mission Control, surrounded by business‑dressed officials, large screens showing a world map, data graph and Earth from space, formal serious atmosphere, high contrast, cinematic lighting, ultra‑sharp 8K detail, 4:3
Tech-stack reverse-engineer
Model: Midjourney v5 – the soft B&W grain, cinematic lighting and typical diffusion blur of text on screens point to Midjourney’s recent photorealistic version
Settings: aspect ratio 16:9 (or 3:2), stylize 100, quality 2, upscaler standard (U), possible NASA/photorealism LoRA
Tells: blurred/indistinct lettering on the control screens, inconsistent focus (president sharp, background slightly soft), slight halo/ghosting around facial edges, uneven grain pattern typical of diffusion, occasional "swirl" artifacts near screen borders, mismatched lighting angles on people vs. room, subtle color channel noise despite black‑and‑white conversion
A realistic high‑resolution mockup of a formal press release: deep blue background, white rectangular text box centered, black serif typeface, header indicating President of Venezuela thanking the United States, corporate official tone, clean minimal design, soft even lighting, 4k, 16:9 aspect
Tech-stack reverse-engineer
Model: Stable Diffusion XL (SDXL) – typical for realistic document mock‑ups; the rendering shows the characteristic diffusion blur and font inconsistencies of SDXL text generation
Settings: aspect ratio ~4:3 (1024x768), low stylize (≈0), high CFG (~12‑15), default SDXL sampler (Euler a), upscaled with ESRGAN/Real‑ESRGAN, using a ‘digital document’ LoRA or style prompt
Tells: fuzzy white‑box edges, slight blue‑background gradient bleed, black text that is not perfectly crisp (minor warping, inconsistent kerning, occasional missing letters or extra spacing), uneven line spacing, tiny artifacts around punctuation, and a subtle halo around the text block – all hallmark diffusion artifacts in generated document images
A collage of realistic news website screenshots showing headlines about Trump's endorsement of Darline Graham in the South Carolina primary, featuring CNN, The Hill, Politico, NPR, The Guardian, The New York Times, The Washington Post, Christian Science Monitor, and Washington Examiner, crisp UI, bright lighting, high detail, 16:9
Tech-stack reverse-engineer
Model: DALL·E 3 – text‑heavy collage prompts often produce garbled logos and article snippets
Settings: aspect ratio ~16:9, low stylize, high quality, default upscaler, no specific LoRA (generic news‑screenshot style)
Tells: blurred or misspelled outlet logos (e.g., "CNN" with extra letters), inconsistent fonts across articles, uneven text spacing, low‑resolution watermark artifacts, mismatched column layouts, faint edge halos around each screenshot, and occasional nonsensical headline wording
Digital collage of 2026 South Carolina Senate runoff news headlines and social media posts, highlighting Donald Trump's endorsement of Darline Graham, crisp newspaper fonts, Twitter‑style bubbles, red‑white‑blue palette, modern magazine layout, high‑resolution, dramatic lighting, photorealistic, 16:9
Tech-stack reverse-engineer
Model: Stable Diffusion XL (text‑to‑image with inpainting) – the mix of crisp layout and garbled, blurry headline text is typical of SDXL’s handling of dense textual scenes
Settings: aspect ratio ~16:9, low stylize (≈250), high quality (≈2), upscaler: 2x ESRGAN, LoRA: ‘text_inpaint_v2’ or similar news‑layout reference
Tells: blurred/warped headline fonts, inconsistent typefaces across sources, missing punctuation or letters, slight halo/edge artifacts around text blocks, mismatched background textures behind the text, occasional repetition or duplication of words, and faint AI watermark signatures in the corner
Create a 2‑column collage of six realistic Twitter posts dated 8.26.2026 about Trump endorsement success. Each tweet shows its image (flag‑painted man, news broadcast screenshots, woman at podium, handshake, etc.) and caption text, Twitter UI style, crisp detail, vibrant lighting, 16:9, ultra‑high‑resolution.
Tech-stack reverse-engineer
Model: Stable Diffusion XL (SDXL) – the mix of realistic portrait‑style subjects, generic news‑broadcast screenshots, and subtle texture mismatches are characteristic of SDXL inpainting and collage workflows
Settings: aspect ratio ~16:9 per tweet tile, low stylize (≈250), high quality (≈2), upscaler: R‑ESRGAN 4x, LoRA: photorealistic‑faces + news‑broadcast‑template, seed varied per tile
Tells: inconsistent lighting and shadow direction between adjacent tweets, slightly blurry or haloed text edges on screenshots, odd facial asymmetry on the flag‑painted man, repetitive generic newsroom background patterns, mismatched pixel density (some tiles look 2‑K, others 720p), occasional ‘floating’ compression artifacts around text boxes, and unnatural eye reflections typical of SDXL generated portraits
Photorealistic flat‑lay collage of newspaper and website screenshots and tweets about the 2026 South Carolina Senate primary – Fox News article on Trump endorsing Darline Graham, Newsmax piece, MAGA Voice post, tweets from DeVory Darkins and Hugh Hewitt, showing candidates Darline Graham, Ralph Norman, Mike Mazzei; bright lighting, high detail, 16:9
Tech-stack reverse-engineer
Model: DALL·E 3 – it excels at generating realistic UI mock‑ups but often leaves subtle text glitches and logo mis‑placements
Settings: aspect ratio 16:9, low stylize (realistic), high quality, upscaler default, no specific LoRA (generic photo‑realistic UI prompt)
Tells: slightly blurry or pixelated text in tweets and article headers, inconsistent font sizes between sources, mis‑aligned logos (Fox News, Newsmax, MAGA Voice), occasional missing letters or swapped characters, uneven spacing in social‑media handles, faint halo around UI elements, minor JPEG‑style compression artifacts around edges
Photorealistic cinematic portrait of a blonde‑haired man in a black suit, white shirt, tie and American flag lapel pin, wearing sunglasses, gripping a large black gun pointed down, standing before a burning city skyline with smoke, flames, debris, distant mountains under a blue sky; dramatic lighting, high detail, epic, 8k, 16:9
Tech-stack reverse-engineer
Model: Midjourney V5 – the composition, dramatic lighting and typical MJ over‑sharp subject with softer, slightly blurred background are characteristic
Settings: Aspect 16:9, Stylize 750, Quality 2, Upscaler Standard, no explicit LoRA (default MJ photorealistic style)
Tells: unnatural lighting on the suit vs. fire (high‑key face, low‑key background), slightly blurred edges on the flames and smoke, overly smooth skin and sunglasses glare, inconsistent perspective between the man’s hands/gun and the distant cityscape, minor distortion in the American flag pin detail, subtle ghosting around the gun barrel
Educational infographic showing U.S. Territories flags, title "U.S. Territories" with small map of the Strait of Hormuz labeled NEW above, flags arranged 5 top, 6 bottom on light blue background, clean simple font, flat vector style, crisp high detail, 16:9
Tech-stack reverse-engineer
Model: Stable Diffusion (grid‑layout LoRA) – typical SD artifacts appear in rendered text and uniform flat flags
Settings: aspect ratio ~4:3 (e.g., 1024x768), low stylize (stylize 0‑10), high CFG ~12, upscaler Real‑ESRGAN, LoRA: flag_grid / flat‑illustration, quality high
Tells: title and map text show blurred, missing or mis‑aligned letters ('U.S. Territories', 'NEW'), flags have slightly uneven borders and inconsistent proportions, color banding on solid backgrounds, subtle tiling/ diffusion noise in the light‑blue backdrop, and a lack of realistic lighting or shadows
President Trump and Dolly Parton standing side by side in front of the White House, watercolor painting, soft dreamy, formal dignified, bright daylight, high detail, 4k, --ar 3:2
Tech-stack reverse-engineer
Model: DALL·E 3 (OpenAI) – renders clear, legible text and soft watercolor style well
Settings: aspect ratio ~3:2, default quality, style prompt includes "watercolor painting", no upscaler, no LoRA
Tells: crisp, readable title text at top (AI models like Midjourney usually garble text), smooth brush‑stroke texture with uniform watercolor wash, slight over‑blending of facial features, simplified architectural details of the White House, muted color palette with pastel pink and muted suit tones, lack of fine‑grain texture typical of real watercolor paper
Ultra‑realistic political collage: Donald Trump in dark suit, stern, centered, surrounded by waving American flags, fighter jets, navy ships, bomb craters; signs reading “MISSION NOT ACCOMPLISHED 2001‑2021” and “MISSION ACCOMPLISHED 2026”; George W. Bush portrait with flames, Ayatollah Khamenei portrait; cinematic dramatic lighting, high detail, 16:9
Tech-stack reverse-engineer
Model: Midjourney V5 – collage‑style prompts often produce the mixed‑subject layout described
Settings: aspect ratio 16:9, stylize 250, quality 2, default upscaler (U2), no explicit LoRA but likely using the default photorealistic style
Tells: hard edge halos around cut‑out figures, mismatched lighting and shadows between Trump, flags, aircraft and ships, inconsistent depth‑of‑field (some elements ultra‑sharp, others soft), slightly blurred or pixelated text on the signs, JPEG‑style ringing around high‑contrast borders, and a subtle “grid‑line” artifact where the separate source images were stitched together
two side‑by‑side portrait panels of a young boy wearing a red‑yellow plaid suit jacket, white shirt and green tie on a light blue background; left portrait neutral expression, right portrait cartoonish big smile; bold black "ME WORRY?" text at bottom; meme style, crisp digital illustration, high resolution, 1:1
Tech-stack reverse-engineer
Model: Stable Diffusion 1.5 – typical for quick portrait memes with duplicated composition
Settings: aspect ratio ~1024x512 (1:2), low stylize (‑‑stylize 0), high quality (‑‑steps 50‑70, CFG 7‑9), upscaler ESRGAN/R‑ESRGAN, LoRA: portrait_v1 + cartoon_smile
Tells: seam visible where the two copies meet, slightly mismatched lighting/shadows between the two faces, overly smooth skin with loss of fine pores, color‑bleeding in the plaid jacket, crisp black text "ME WORRY?" that looks overlaid rather than rendered, occasional blurry edge around the tie and background banding typical of SD in‑painting
Twitter screenshot, tweet from @johnnymaga (parody of @realDonaldTrump), black text on white background, simple UI, profile picture a cartoon character with red hat, quote about Trump ranking first among Republican admirers, clean minimal design, high detail, realistic, 16:9
Tech-stack reverse-engineer
Model: DALL·E 3 – the text‑heavy UI often shows slightly fuzzy fonts and misplaced UI elements typical of DALL·E’s rendering of social‑media screenshots
Settings: aspect ratio ~16:9 (horizontal tweet), low stylize/standard quality, no upscaler, Twitter UI LoRA (e.g., "twitter_ui" or "tweet_screenshot")
Tells: blurred or slightly uneven font rendering, missing/incorrect verification badge, profile picture circle with pixelation, timestamps and separators mis‑spaced, rounded‑corner tweet box not perfectly smooth, subtle shadow inconsistencies, occasional missing or extra characters in the text
From passenger seat, view through windshield of night drive, car interior with red air freshener, tropical road lined with palm trees and streetlights, white truck ahead, dark cloudy sky, large blue sign with white text and gold eagle reading 'WELCOME TO PRESIDENT DONALD J. TRUMP INTERNATIONAL AIRPORT', cinematic, hyper‑realistic, 8K, shallow depth of field, neon glow --ar 16:9
Tech-stack reverse-engineer
Model: Midjourney v5.2 – the composition, depth‑of‑field and slightly stylized lighting match MJ’s photorealistic car‑interior outputs
Settings: aspect 16:9, stylize 250, quality 2, upscaler default (MJ upscaler), LoRA: photorealistic_car_interiors
Tells: glass reflections are blurry or missing on the windshield, the sign’s text shows slight font inconsistencies and edge artifacts, palm‑tree silhouettes lack fine leaf detail, lighting on the dashboard doesn’t match the night exterior, and the red air‑freshener has a plastic‑like edge halo typical of MJ render artifacts
