GPT Image 2 is OpenAI's new generation image model (codename "duct-tape"). It supports pixel-level multilingual text rendering, strong consistency, and production-grade illustration quality.

Where do these prompts come from?

They are mainly curated from public Twitter/X posts. If you are an original author and do not want your prompt listed, contact us and we will remove it promptly.

How can I use these prompts?

For now, copy a prompt and use it with the GPT Image 2 model inside Felo. We will improve the integration later so you can run it in one click.

GPT Image 2 Prompts - Pixel-level Text Rendering & Production-grade AI Images

AuthorSRKDAN

154 days ago

Seamless Jacket Design

A creative fashion prompt for generating a futuristic jacket without any zippers, buttons, or visible fasteners.

Prompt

A {argument name="clothing item" default="jacket"} with no zip, no buttons, no fastening — closes perfectly.

Authorstudioあぽろん

154 days ago

Anime Girl Smug Peace Sign Portrait

Generates an anime-style portrait of a girl making a peace sign with customizable hair, eyes, and background colors.

Prompt

An anime-style illustration of a young girl with {argument name="hair color" default="silver"} hair styled in long twin tails secured with black ribbons, and {argument name="eye color" default="pink"} eyes. She has a {argument name="expression" default="smug smile with a slight blush"} and is looking directly at the viewer. She is posing dynamically, leaning forward slightly with one hand making a peace sign near her cheek and the other hand resting on her hip. She is wearing a white long-sleeved blouse featuring ruffles down the center and at the cuffs, a black ribbon tied at the collar, and a black high-waisted skirt with a ruffled hem and corset-style lacing. The background is a solid {argument name="background color" default="vibrant pink"} with a subtle vignette effect. The art style is high-quality, detailed anime illustration with soft lighting and glossy hair highlights.

Authorらけしで / Mizmic Studio

154 days ago

Anime Girl in Space Elevator

Generates an anime-style illustration of a character looking up at a massive space structure from inside a glass elevator.

Prompt

A high-angle, anime-style illustration of a girl with {argument name="hair color" default="long dark hair with purple tips"} standing inside a futuristic glass elevator, looking up at a massive space elevator structure. She is wearing a {argument name="outfit" default="sailor uniform with a pleated skirt and thigh-high socks"}. The elevator cab has metallic railings, glowing cyan accents, and a digital display showing an upward arrow and the text "{argument name="altitude text" default="005829 km"}". Through the large glass ceiling and windows, a colossal, intricately detailed space station with a central pillar and a large ring structure is visible against a backdrop of a starry sky and the curvature of {argument name="planet" default="Earth"} with clouds below. The lighting is dramatic, emphasizing the vast scale of the space structure and the atmospheric perspective.

AuthorNeotype

155 days ago

Historical Social Media Feed Mockup

A creative prompt to generate a fictional Twitter-like (X) social media feed showing a historical argument between General Yi Seong-gye and General Choe Yeong.

Prompt

Please create an X (Twitter) page for {argument name="인물" default="King Taejo Yi Seong-gye"} ({argument name="상황" default="just before the Wihwado Retreat"} - containing posts dissing each other with {argument name="상대" default="General Choe Yeong"}).

Author結パパ

155 days ago

Anime Noodle Advertisement Banner

Generates a vibrant, text-heavy Japanese promotional banner featuring an anime character, a product mockup, and a steaming bowl of noodles.

Prompt

{ "type": "Japanese promotional banner advertisement", "background": "vibrant red with subtle sakura cherry blossom patterns and gold confetti", "character": { "style": "anime", "appearance": "young woman with long black braided hair, wearing a fox mask on the side of her head", "outfit": "{argument name=\"character outfit\" default=\"red and white miko shrine maiden dress with decorative knots\"}", "pose": "winking, smiling energetically, pointing at the product cup" }, "product": { "description": "instant noodle cup mockup", "design": "red and white packaging with 'COP NEEDLE' and '醤油味' text" }, "food_shot": { "description": "large steaming bowl of ramen noodles", "toppings": ["3 shrimp", "cubed meat", "half boiled egg", "scallions", "yellow egg bits"] }, "typography_and_layout": { "text_elements_count": 8, "elements": [ { "position": "top-left starburst", "text": "{argument name=\"top left burst text\" default=\"新発売!\"}" }, { "position": "top-center huge 3D title", "text": "{argument name=\"main product name\" default=\"コップヌードル\"}" }, { "position": "below main title pill", "text": "醤油味" }, { "position": "center yellow jagged burst", "text": "お湯を注いで 3分!" }, { "position": "mid-right floating text", "text": "{argument name=\"catchphrase text\" default=\"だし香る、やさしい味わい！コクうまスープ！\"}" }, { "position": "bottom-left speech bubble", "text": "巫女ちゃんおすすめ!" }, { "position": "bottom-center text", "text": "小腹がすいたらこれ!" }, { "position": "bottom-right red price box", "text": "{argument name=\"price text\" default=\"198円\"}" } ] } }

Authorらけしで / Mizmic Studio

155 days ago

Fantasy Manga Character Introduction Page

Generates a 6-panel manga page introducing a party of 5 fantasy characters with dialogue and name boxes.

Prompt

{ "type": "manga page", "style": "{argument name=\"art style\" default=\"black and white manga screentone\"}", "character_count": 5, "panel_count": 6, "layout": { "panels": [ { "position": "top, full width", "character": "Beatrice, long blonde hair, dark hooded cloak, pointing forward", "text": ["We will challenge the legendary ruins with knowledge and the power of our friends! Are you ready?", "Let's go!", "Future Scholar {argument name=\"character 1 name\" default=\"Beatrice\"}"] }, { "position": "middle left", "character": "Nova, long black hair, elf ears, black dress, arms crossed", "text": ["I hate trouble, but... I'll tag along.", "Demon Lord's Daughter {argument name=\"character 2 name\" default=\"Nova\"}"] }, { "position": "middle right", "character": "Crim, short hair, armored corset, holding sword", "text": ["Of course! We'll defeat the monsters and win together!", "Hero {argument name=\"character 3 name\" default=\"Crim\"}"] }, { "position": "bottom left, upper", "character": "Lilia, witch hat, holding staff", "text": ["So exciting!! I want to be useful! I'll learn a lot and grow!", "Fledgling Adventurer Lilia"] }, { "position": "bottom right, tall vertical", "character": "Camilla, dragon horns, wings, tail, black armor, large sword", "text": ["I will use my power if needed.", "Dragon Girl Camilla"] }, { "position": "bottom left, lower", "description": "Group shot of all 5 characters standing in a landscape", "text": ["Let's set off!", "*step*", "The first step towards the legend with these unique companions begins now!"] } ], "footer": "{argument name=\"ending text\" default=\"The End\"}" } }

Authorかず｜SE×AI副業

155 days ago

Photorealistic Portrait of a Young Woman

Generates a highly detailed, photorealistic portrait of a young woman with customizable features, clothing, and background.

Prompt

A photorealistic, high-quality portrait of a beautiful young {argument name="nationality" default="Japanese"} woman. She has {argument name="hair style" default="long black wavy hair with bangs"}, dark brown eyes, fair skin, and a {argument name="expression" default="gentle smile"} while looking directly at the camera. She is wearing a {argument name="clothing" default="simple white blouse"}. The image features soft, flattering portrait lighting and a shallow depth of field, set against a {argument name="background" default="blurred warm indoor setting"} with beautiful bokeh. Shot on an 85mm lens, highly detailed.

Authorかず｜SE×AI副業

155 days ago

Elegant Photorealistic Female Portrait

Generates a highly detailed, photorealistic portrait of a woman in an elegant setting with customizable features and attire.

Prompt

A photorealistic portrait of a {argument name="subject" default="beautiful young Asian woman"} with {argument name="hair style" default="long dark wavy hair with side-swept bangs"}. She wears an {argument name="outfit" default="elegant black off-the-shoulder satin dress"} and pearl stud earrings, smiling gently at the camera. The background is a {argument name="background setting" default="blurred luxurious interior with warm bokeh lights"}. Cinematic lighting, highly detailed.

AuthorEARLの医学&AIノート

155 days ago

Photorealistic Neon Tokyo Street Scene

Generates a highly detailed, photorealistic night view of a bustling Japanese city street with customizable neon signs and billboards.

Prompt

A photorealistic night street photography shot of a bustling {argument name="city location" default="Tokyo"} street, heavily illuminated by densely packed vertical neon signs and billboards. On the left, a bright pink sign prominently reads "{argument name="pink sign text" default="カラオケ"}" next to a white sign reading "牛角焼肉". On the right, prominent signs include a yellow one reading "{argument name="yellow sign text" default="まんが喫茶 4F"}", a white one reading "東南本館 2F", and a red one reading "{argument name="red sign text" default="中華食堂一番館 B1"}". In the center distance, a large blue billboard displays "{argument name="billboard text" default="open TOKYO"}" above a digital screen showing a person with the text "YUNIKA VISION". The street level is crowded with the backs of pedestrians' heads walking through the vibrant, glowing urban canyon. The lighting is high contrast with rich neon colors reflecting off the surroundings.

AuthorKouji Fujii｜AI × Business

155 days ago

Photorealistic Shibuya Street Selfie

Generates a photorealistic selfie of a young woman on a Japanese city street with customizable text on signs and buildings.

Prompt

A photorealistic selfie of a young Japanese woman standing on a sunny city street in {argument name="location" default="Shibuya, Japan"}. She has a soft smile, fair skin, light makeup, and {argument name="hair style" default="long wavy brown hair with see-through bangs"}. She is wearing a {argument name="clothing" default="white textured blouse"} with a delicate necklace and a beige shoulder bag strap visible. The background features a bustling urban environment with a crosswalk, pedestrians on the sidewalk, and a large white commercial building displaying the text {argument name="building text" default="MODI"} at the top along with a digital billboard. On the left side of the frame, a vertical street sign clearly displays the text {argument name="street sign text" default="渋谷区神南一丁目 8"}. The image has a bright, natural daylight aesthetic typical of a casual social media photo.

AuthorKouji Fujii｜AI × Business

155 days ago

Handwritten Japanese Study Notes

A photorealistic top-down view of an open notebook with handwritten Japanese study notes and a pencil, useful for educational or aesthetic mockups.

Prompt

A top-down photorealistic image of an open lined notebook resting on a smooth wooden desk. A dark green HB pencil with a silver ferrule and white eraser lies diagonally on the desk to the right of the notebook. The notebook pages are filled with neat, handwritten text in pencil, resembling a student's study notes. On the left page, the top left corner reads {argument name="date" default="2024年5月20日(月)"} above {argument name="subject" default="国語「枕草子」"}. Below this is a rectangular boxed heading reading {argument name="left page heading" default="清少納言のものの見方"}, followed by exactly three bulleted sentences: one about valuing the changes in nature and seasons, one about finding beauty and amusement in small everyday things, and one about being honest with one's senses while having a sense of humor. Further down is a cloud-shaped border containing a heading translated as 'Words that left an impression', followed by a famous classical quote about spring mornings and a brief explanation below it marked with an arrow. On the right page, the top section has the heading {argument name="right page heading" default="〈現代とのちがい〉"} with exactly two bulleted points: one noting that the perception of seasons is much more delicate than today, and another stating that the perspective is sensory and poetic. This is followed by an arrow pointing to a concluding thought about wanting to cherish this sensibility. Below this is another cloud-shaped border containing a heading translated as 'Summary' and a short paragraph about how the 1000-year-old text still resonates today and the importance of carefully observing nature. At the bottom right of the page, there is a small, simple line drawing of a leafy plant stem next to a scalloped thought bubble containing the text {argument name="final thought" default="感性をみがくって、素敵なことだなあ。"}. The lighting is soft and natural, highlighting the texture of the paper and the graphite pencil marks.

AuthorScott Muller

155 days ago

Paranormal Trail Camera Encounter

Generates a spooky, infrared trail camera image featuring wildlife, a mysterious glowing orb, and realistic camera data overlays.

Prompt

A black and white infrared trail camera photo taken at night in a dense coniferous forest. In the foreground on the left, a large black bear is walking past, appearing slightly motion-blurred. In the center of the clearing, there is a perfectly circular, dark, barren patch of dirt on the forest floor. Hovering just above the ground at the far edge of this dark circle is a bright, glowing, spherical orb of light. The image features a grainy night vision aesthetic with visible underbrush, pine tree trunks, and a tree stump in the bottom right corner. A black data bar runs across the bottom edge containing white text: on the far left, an oval logo reading "{argument name="camera brand" default="RECONYX"}", followed by the timestamp "{argument name="timestamp" default="03:14:22 11-OCT-2023"}", then "{argument name="temperature" default="TEMP: 4°C"}", and on the far right, "{argument name="trigger info" default="TRIGGER: MOTION DUAL TRIGGER"}".

AuthorScott Muller

155 days ago

Satellite Anomaly Interface Mockup

Generates a realistic web interface for satellite imagery featuring a mysterious geometric anomaly, timeline thumbnails, and UI controls.

Prompt

{ "type": "satellite imagery web interface mockup", "main_display": { "imagery": "greyscale satellite view of ocean clouds featuring a {argument name=\"anomaly shape\" default=\"massive, perfectly geometric dark triangle\"} in the center", "annotation": "red handwritten text '{argument name=\"annotation text\" default=\"Natural formation???\"}' with a red arrow pointing to the anomaly", "overlay": "faint blue latitude and longitude grid lines with coordinate labels like 40N, 30N, 160W" }, "ui_layout": { "top_header": { "left_section": "NOAA logo, text '{argument name=\"satellite name\" default=\"GOES-18 ABI CONUS - Sector View\"}', '{argument name=\"timestamp\" default=\"2024-03-15 14:32:20 UTC\"}', and 'Band 2 (0.64µm) - Visible Light'", "right_section": "4 navigation icons with labels: Home, Share, Help, Settings" }, "left_sidebar": { "count": 7, "items": ["Layers", "Products", "Basemap", "Measure", "Location", "Legend", "Download"] }, "bottom_panel": { "timeline": { "controls": "Play button, left/right arrows, text '6 Scenes', '6 of 6'", "thumbnails": { "count": 6, "description": "sequential satellite images showing the anomaly's progression, with hourly timestamps from 08:32 UTC to 14:32 UTC" } }, "data_readout": { "coordinates": "{argument name=\"coordinate text\" default=\"Lat/Lon: 28.12°N, 162.08°W\"}", "elevation": "Elevation: 0.5°", "scale": "Distance Scale bar from 0 to 1000 km" } } } }

AuthorScott Muller

155 days ago

Mars Rover Discovery Photo

Generates a photorealistic, raw space agency style image of a Mars rover discovering alien artifacts, complete with a file name overlay.

Prompt

A photorealistic, raw space agency photo of a {argument name="setting" default="Martian landscape"} with red dirt, scattered rocks, and a hazy orange sky. In the middle ground, {argument name="objects" default="two dark stone tablets covered in ancient carved hieroglyphs"} are jutting out of the ground at a slight angle. In the immediate foreground, the distinct {argument name="foreground element" default="shadow of a Mars rover"} is cast onto the rocky terrain, indicating the perspective of the photo. Across the top edge of the image, there is a plain white text overlay resembling a raw file name reading "{argument name="overlay text" default="ZCAM_0847_0734291854_678RAD_N0412172ZCAM08050_1100LUJ01.png"}". The lighting is harsh and realistic, mimicking actual extraterrestrial surface photography.

AuthorScott Muller

155 days ago

Google Street View Anomaly Screen Photo

Generates a realistic photograph of a computer screen displaying a Google Street View interface featuring a surreal or out-of-place architectural anomaly.

Prompt

{ "type": "photograph of a computer monitor displaying a Google Maps Street View interface", "scene": { "environment": "{argument name=\"environment\" default=\"barren desert landscape under a clear blue sky\"}", "subject": "{argument name=\"subject\" default=\"massive ancient stone wall with perfectly fitted cyclopean masonry blocking a dirt road, with a rectangular doorway cut through the center\"}", "foreground": "shadow of a Google Street View car with a 360-degree camera mounted on top cast onto the dirt road" }, "ui_elements": { "browser_bar": "URL bar at the top showing google.com/maps with latitude and longitude coordinates", "top_left_panel": "black semi-transparent info box with text 'Unnamed Road', '{argument name=\"region\" default=\"Atacama Region\"}', and 'Street View - Sept 2022'", "bottom_left_panel": "minimap inset with a yellow Pegman icon", "bottom_right_panel": "minimap inset showing a blue route line, Pegman, zoom controls, and text '{argument name=\"distance text\" default=\"340 km from nearest town\"}'", "bottom_bar": "Google logo in the bottom center, compass icon on the right, and copyright/capture date text along the bottom edge" }, "style_details": "visible screen glare and slight monitor distortion to simulate a photograph taken of a physical screen with a smartphone" }

AuthorMatthew Berman

155 days ago

Intimidating Press Conference Meme

A photorealistic satirical image of a superhero intimidating a tech CEO at a White House press briefing.

Prompt

A photorealistic, cinematic shot set in the {argument name="location" default="White House press briefing room"}. At the podium sits {argument name="seated character" default="Sam Altman"}, wearing a dark suit and blue tie, looking visibly uncomfortable and nervous, staring blankly ahead. Standing directly behind him is {argument name="standing character" default="Homelander from The Boys"}, wearing his signature blue superhero suit with red gloves and gold eagle shoulder pads. The standing character has one red-gloved hand resting heavily and intimidatingly on the seated character's shoulder, while using his other hand to drink from a {argument name="beverage" default="glass bottle of milk"} with a confident, slightly unhinged smirk. In the background, there is a US flag on the left and a blue oval White House seal on the wall behind them. The podium has a sign reading "THE WHITE HOUSE WASHINGTON".

AuthorAditya ⚡Rao

155 days ago

Dark Fantasy Game Event UI Mockup

Generates a game event landing page featuring impasto-style combat artwork and localized UI panels for missions and rewards.

Prompt

{ "type": "game event landing page UI mockup", "art_style": "{argument name=\"art style\" default=\"heavy impasto oil painting, dark fantasy\"}", "background_artwork": { "description": "dynamic combat scene with swirling dark and red energy", "character_count": 4, "characters": [ "top: white-haired man casting blue spell", "middle-left: dark-haired man summoning shadow energy", "right: pink-haired man throwing red energy punch", "bottom-left: brown-haired woman wielding hammer" ] }, "ui_layout": { "top_left": { "type": "event header", "elements": [ {"small_text": "{argument name=\"subtitle\" default=\"League of Legends X Jujutsu Kaisen\"}"}, {"large_title": "{argument name=\"main title\" default=\"Sorcerer's Showdown\"}"}, {"date_text": "{argument name=\"event date\" default=\"Event Period: 2024.07.18 - 2024.08.28\"}"}, {"description": "Jujutsu Kaisen has descended upon Summoner's Rift. Complete missions and earn limited rewards!"}, {"button": "{argument name=\"button text\" default=\"Participate in Event\"}"} ] }, "bottom_left": { "type": "mission panel", "header": "Daily Missions 0/8", "mission_count": 2, "missions": [ {"text": "Play 1 game in the Rift", "progress": "0/1", "reward": "100"}, {"text": "Participate in 15 enemy champion takedowns", "progress": "0/15", "reward": "200"} ] }, "bottom_right": { "type": "rewards panel", "labels": ["Jujutsu Kaisen Pass", "Limited Skin Rewards"], "icon_count": 4, "icons": [ "blindfolded man", "pink-haired boy", "brown-haired girl", "dark monster" ] } } }

AuthorAditya ⚡Rao

155 days ago

Live Stream Interface Mockup

Generates a realistic live streaming interface featuring a host, live chat overlay, topic cards, and video player controls.

Prompt

{ "type": "live stream interface mockup", "subject": { "description": "{argument name=\"subject appearance\" default=\"young Korean woman, long dark wavy hair, light blue button-up shirt\"}", "accessories": "silver cross necklace, black broadcast microphone on boom arm", "pose": "speaking to camera, sitting in grey ergonomic chair" }, "background": "{argument name=\"background vibe\" default=\"cozy room with warm lighting, shelf with YouTube play buttons\"}", "layout": { "top_bar": { "title": "{argument name=\"stream title\" default=\"What will the AGI era bring us? | Technology, Jobs, Meaning of Human\"}", "badge": "LIVE" }, "left_chat": { "count": 7, "messages": [ {"user": "Luna_1023", "text": "Really looking forward to today's topic!"}, {"user": "Future_Explorer", "text": "If AGI becomes smarter than humans, what kind of beings will we become?"}, {"user": "CodenameA", "text": "Can you also talk about jobs? I'm so anxious lately..."}, {"user": "StarlightLove", "text": "Let's think about it together while listening to Luna's voice"}, {"user": "AI_Tamer", "text": "Is AGI a tool or a companion?"}, {"user": "MeowPunch", "text": "Wow! Exciting from the start!"}, {"user": "PhilosophizingPrince", "text": "'Meaning of Human'... Expecting a deep conversation already"} ] }, "bottom_left_card": { "header": "TODAY'S TOPIC", "main_text": "{argument name=\"topic title\" default=\"The AGI Era, Where Are We Going?\"}", "sub_text": "Technology · Jobs · Meaning of Human" }, "bottom_controls": { "elements": ["red progress bar", "play icon", "volume icon", "LIVE", "settings icon", "fullscreen icon"] } } }