APIMODELSAPIMODELS

25+ GPT Image 2.5 Product & Advertising Prompts

25+ prompts. Copy-ready GPT Image 2.5 prompts, each with the image it produced and a link to where it came from. Text-to-image, precise edits, infographics, UI mockups, sprite sheets and product shots.

← All 238 GPT Image 2.5 prompts

Last updated: · GPT Image 2.5 API — pricing & playground · API docs

TL;DR

GPT Image 2.5 is the image model OpenAI shipped on 9 September 2026, in two API variants: Flare, which matches GPT Image 2 on quality and editing at roughly half the latency, and Sunburst, which trades time for precision on demanding work. Both render text cleanly, generate natively at 1K, 2K and 4K, and can return a transparent background. The headline change is controllable editing: you name what must not move, and it does not move. This page collects prompts that people actually ran, with the resulting image and a link to the source. On apimodels.app both variants are live, priced by resolution and quality — $0.01–$0.08 at 1K and $0.020–$0.15 at 2K for either, and at 4K $0.02–$0.3 on Flare or $0.02–$0.47 on Sunburst — billed only on success.

Source: OpenAI image prompting guide (the model vendor’s official site)

Shampoo highway billboard

Shampoo highway billboard

Create a realistic billboard mockup of the shampoo on a highway scene during sunset. Billboard text (EXACT, verbatim, no extra characters): "Fresh and clean" Typography: bold sans-serif, high contrast, centered, clean kerning. Ensure text appears once and is perfectly legible. No watermarks, no logos.

by OpenAISource
Retro toy plane collector packaging

Retro toy plane collector packaging

Create a collectible action figure of a vintage-style toy propeller airplane with rounded wings, a front-mounted spinning propeller, slightly worn paint edges, classic childhood proportions, designed as a nostalgic holiday collectible, in blister packaging. Concept: A nostalgic holiday collectible inspired by the simple toy airplanes children used to play with during winter holidays. Evokes warmth, imagination, and childhood wonder. Style: Premium toy photography, realistic plastic and painted metal textures, studio lighting, shallow depth of field, sharp label printing, high-end retail presentation. Constraints: - Original design only - No trademarks - No watermarks - No logos Include ONLY this packaging text (verbatim): "Christmas Memories Edition"

by OpenAISource
Person holding the product, styled after a reference

Person holding the product, styled after a reference

Using this JSON as reference, generate a person holding my product.

by @Mho_23Sourcevia MagicCreator
UGC first frame: phone showing the app

UGC first frame: phone showing the app

create a HQ hyper-realistic photo of a gen Z girl holding a phone up with this screen showing an app exactly like the reference photo in a UGC style video first frame

by @eptwtsSourcevia MagicCreator
Dense frame: 1994 infomercial freeze

Dense frame: 1994 infomercial freeze

a frame grab from a 1994 infomercial, standard definition video with scanlines. A host in a teal blazer, a battered saucer on a rotating plinth with Antarctic ice frozen around one leg, a price starburst, a phone number, and three lines of legal small print.

Real object plus hand-drawn doodle ad

Real object plus hand-drawn doodle ad

一则简约且富有创意的商品广告,纯白背景。一台真实的复古胶片相机与手绘黑色墨水涂鸦相结合,线条松散而俏皮:涂鸦小人正用小推车搬运胶卷、踮脚给相机喂一颗齿轮形状的糖,一只涂鸦小鸟停在闪光灯上梳理羽毛。画面上方加入粗体黑色广告文案「拍下生活的每一格」,底部放置简洁的「菲林社 FILM CLUB」文字标志与一个小太阳图形。视觉效果简洁、有趣、高对比度,构思巧妙。

Sourcevia LaplaceYoung
Chrome emoji badge

Chrome emoji badge

高精度 3D 渲染:一枚金属质感徽章,主体是立体圆润的⭐星星 emoji 造型,超光滑镀铬质感,固定在竖直的商品展示卡上,风格化的未来主义设计,柔和反光与干净阴影。纸质卡片顶部中央带一个冲切欧式挂孔,徽章上方是醒目标题「STAR POWER」,下方配趣味标语「Shine on, you crazy pixel!」。背景柔和灰色,柔光摄影棚灯光,极简风格。

Sourcevia LaplaceYoung
Animal silicone wrist rest

Animal silicone wrist rest

创建图片:一个可爱 Q 版的硅胶护腕托,外形基于🐼熊猫,采用柔软食品级硅胶材质,表面亲肤哑光质感,内部填充慢回弹棉。拟人化卡通风格,表情生动,双手张开趴在桌面上,呈现拥抱手腕的姿势,整体造型圆润软萌,黑白熊猫配色,治愈可爱,适合办公使用。白色纯色背景,柔和布光,产品摄影风格,45° 俯视视角,高清细节突出硅胶质感与舒适功能。

Sourcevia LaplaceYoung
Bookshelf shaped like a logo

Bookshelf shaped like a logo

拍摄一张现代书架的照片,造型灵感来源于无穷符号 ∞ 的形状:书架由流畅、互相连接的曲线构成,形成多个大小不一的分区。材质为光滑的哑光黑色金属,曲线内部设木质层板,柔和暖色的 LED 灯带勾勒内侧曲线轮廓。书架安装在中性色调的墙面上,摆放色彩丰富的书籍、小型绿植和极简风格艺术摆件。整体氛围富有创意、优雅且略带未来感。

Sourcevia LaplaceYoung
Fantasy underwater popsicle

Fantasy underwater popsicle

倾斜的第一人称视角:一只手握着一支超现实冰棒。透明的蓝色冰壳里封存着一个微型水下世界——小潜水员、几条小鱼、漂浮的气泡和翻滚的海浪,一根绿色冰棒棍贯穿中心。冰棒略微融化,一滴水正顺着木棍滑落。背景是柔焦的上海梧桐区街景,高端产品摄影风格,3:4 竖版。

Sourcevia LaplaceYoung
Emoji ice cream bar

Emoji ice cream bar

生成图片:将🍓变成一根奶油雪糕,奶油在雪糕顶部呈丝滑曲线流动,看起来美味可口,45° 悬浮在空中,Q 版 3D 可爱风格,与草莓色系一致的纯色背景,柔和影棚光,1:1。

Sourcevia LaplaceYoung
Emoji tufted rug

Emoji tufted rug

创建一张图像:一块彩色手工簇绒地毯,形状为🦖恐龙 emoji,铺设在简约的木地板上。地毯设计大胆俏皮,柔软蓬松的质感配粗线条细节,绿色为主、奶白点缀。从上方俯拍,自然光照,整体风格略带古怪的 DIY 手作美感——类似手工簇绒艺术地毯,边缘绒线微微翘起。

Sourcevia LaplaceYoung
Creative planter

Creative planter

一张高质量照片:一个可爱的陶瓷柴犬形状花盆,釉面光滑,里面种满生机勃勃的多肉植物——尖刺的十二卷、莲座状的石莲花和精致的白色小花。花盆带一张友好的微笑面孔,放置在柔和的中性背景上,漫射自然光照明,展示细腻的釉面纹理与植物的色彩对比,构图简洁、极具简约风格。

Sourcevia LaplaceYoung
Inflatable emoji cushion

Inflatable emoji cushion

创建一个高分辨率的 3D 渲染图:将🥹设计成一个充气鼓胀的靠垫物体,形状柔软、圆润、充满空气,类似毛绒气球或充气玩具。使用光滑哑光材质,带细微布料折痕和缝线以强化充气效果;整体形态略不规则、柔软塌陷,配柔和阴影与软光照突出体积感与真实感。置于干净简约的浅灰蓝色背景上,风格俏皮而具雕塑感。

Sourcevia LaplaceYoung
ESC keycap diorama

ESC keycap diorama

一个超写实的等距视角 3D 渲染图:一个微型电脑工作空间被装在一个半透明机械键盘键帽内,键帽特别放置在一块哑光机械键盘的 ESC 键上。键帽内部,一个穿着舒适连帽衫的小人坐在人体工学椅上,专注面对发光的电脑屏幕工作;空间布满逼真的微型配件——真实材质台灯、带反光的显示器、微小扬声器格栅、缠绕的电缆和陶瓷杯。键帽底部由土壤、岩石和苔藓构成,照片级材质质感;内部光照模拟温暖的傍晚氛围,与冷色屏幕光形成对比。

Sourcevia LaplaceYoung
Happiness capsule

Happiness capsule

一张产品概念广告图。标题(大字):速效快乐胶囊。主体:一颗上半为咖啡棕色、下半透明的小药丸,印有虚构品牌「MochaCat」的小猫头像 logo,胶囊里塞满烘焙咖啡豆。说明(小字):请在悲伤难过时服用,一日三次,一次两粒。下方是与药丸颜色一致的购买按钮,价格 $9,角落一行极小的字:请遵医嘱酌情购买。干净的米色背景,药丸微距广告摄影风格,柔和布光。

Sourcevia LaplaceYoung
Soft-style 3D advertisement

Soft-style 3D advertisement

一个柔和的 3D 卡通风格雕塑广告:虚构无线耳机品牌「AirPebble」,由光滑的粘土质感与鲜艳柔和的色彩制成。简约等距场景中,鹅卵石形状的充电盒半开,两枚耳机像小石子一样悬浮升起,旁边点缀真的鹅卵石与苔藓。构图简洁,光线柔和,阴影微妙,画面下方清晰显示产品 logo 与三词口号「小石头,大安静」。

Sourcevia LaplaceYoung
Fluffy jack-o'-lantern

Fluffy jack-o'-lantern

将简单平面的矢量图标🎃转化为柔软、立体、毛茸茸的可爱物体:整体被浓密蓬松的橙色绒毛完全覆盖,毛发质感极其真实,带柔和阴影;南瓜灯的锯齿眼睛与微笑由深棕色绒毛精心拼出。物体居中悬浮于干净的浅灰色背景中,轻盈漂浮。整体风格超现实、富有触感和现代感,带来舒适俏皮的视觉感受。摄影棚级灯光,高分辨率渲染,1:1。

Sourcevia LaplaceYoung
Country diorama in a toy box

Country diorama in a toy box

一张超写实的俯拍摄影作品:两只手撑开一个米色纸板盒的盒盖,盒内是 3D 打印的「中国」微缩立体模型——长城蜿蜒、黄土高原梯田、江南水乡、上海天际线、大熊猫栖息的竹林,以及大量微小精细的人物模型。全部采用触感舒适的玩具风格,哑光 3D 打印纹理并带可见的打印层纹。盒盖内侧用大号、色彩鲜艳的凸起塑料字母显示「CHINA」。光线温暖且具电影感,突出纹理与阴影。

Sourcevia LaplaceYoung
Painting figure in a cereal ad

Painting figure in a cereal ad

《大师麦片》广告大片:以《戴珍珠耳环的少女》的画中人物为主角,根据她的气质设计一款燕麦产品(蓝莓、酸奶、燕麦与珍珠大麦的搭配),并生成她作为包装盒封面人物的广告场景。人物保留名画中的蓝黄头巾与珍珠耳环特征,但转为可爱 Q 版 3D、C4D 渲染风格。包装盒印品名「VERMEER 珍珠麦片」与 slogan「每一口都是名画」,盒子摆在极简主义设计台面上,柔和晨光。

Sourcevia LaplaceYoung

Commercial pink berry smoothie photography

在明亮舒适的现代厨房中,创作一张超写实的高端食品摄影作品。左侧的白色圆桌上,垂直摆放着一组 鲜艳浓郁的粉色浆果冰沙,盛装在 三个小型透明玻璃罐叠放 中。冰沙质地浓稠细腻,带有细微的浆果颗粒和光泽感。右侧放置一个带有绿色盖子的透明 有机花生酱 罐,里面装满了金棕色的奶油花生酱。请准确还原产品包装、标签位置、颜色和比例。在罐子旁边自然地散落几颗完整的花生。前景处放置一个装有混合烤花生的小陶瓷碗,以及一个米色陶瓷盘,盘中放着一把盛有满满一勺奶油花生酱的勺子。在冰沙旁点缀几枝精致的绿色香草。背景:整洁的白色地铁砖厨房墙面,侧面照射的柔和自然光,微妙的阴影,营造出温暖舒适的氛围。将食物放置在一个小型木制项目上,下方垫着一张复古印花纸。在左上角添加优雅的手写风格白色文字“Smoothies”,下方配以较小的草书文字“Tingi Kalori”。垂直 9:16 构图,高端商业食品摄影,逼真的质感,自然日光,浅景深,柔和的背景虚化,清晰的产品细节,平衡的构图,温暖的生活方式美学,照片级真实感,高分辨率,无人物。

Sourcevia LaplaceYoung

Luxury watch product ad template

{ "type": "产品广告", "objective": "为奢华腕表创建电商广告", "inputs": { "subject": "黑金机械腕表", "scene": "黑色石质底座", "style": "奢华产品摄影", "palette": "黑色、金色、深灰色" }, "quality_constraints": { "aspect_ratio": "4:5", "composition": "居中放置,顶部留有文案空间" }, "output_requirements": { "usage": "电商广告", "focus": "表盘与金属质感" } }

Sourcevia LaplaceYoung
AdRevival private management storyboard

AdRevival private management storyboard

Create a complete cinematic storyboard and styleframe board for a premium fintech / advertising management SaaS product called “AdRevival Private Management”. The final output should be ONE SINGLE LARGE STORYBOARD BOARD, not one individual slide. Use a dark charcoal-gray infinite-canvas background, like a professional Figma or creative-director presentation board. Arrange approximately 20 polished 16:9 video keyframes in a clean horizontal production storyboard, grouped into clearly separated narrative sections. Add small black rounded section labels above the columns: “Hook” “Intro” “Use case” “CTA” Every frame should have a tiny scene label in its upper-left corner such as “Scene 1a”, “Scene 1b”, “Scene 2a”, etc. VISUAL DIRECTION: Premium, cinematic, minimal, high-end motion-design style. Black and near-black backgrounds, deep midnight blue, subtle violet-blue light, champagne gold and warm yellow accents, white typography. Elegant gradients, soft bloom, subtle grain, faint dot-grid textures, volumetric glow, dramatic vignettes, polished reflections. Clean Swiss / Apple / fintech editorial typography, similar to Helvetica Neue / Inter / SF Pro. Large kinetic typography combined with clean SaaS dashboard UI. Sophisticated but restrained, not flashy. Consistent visual identity across every frame. Professional motion graphics agency quality. No random stock photography. No clutter. All frames should look like finished title cards from the same commercial. STORY / SHOT SEQUENCE: HOOK Scene 1a: Minimal dark blue frame. A small circular AdRevival brand symbol in the center followed by the word “For”. Lots of negative space. Scene 1b: Same composition evolving into the phrase “For 2 years”, with “2 years” subtly highlighted in blue-violet. Scene 1c: Centered “AdRevival” wordmark and circular symbol on a deep blue-black gradient background. Scene 1d: A dramatic monetary milestone frame showing “$500,000,000” in elegant white type, surrounded by curved orbital lines and several glowing social / advertising platform icons, including a TikTok-like icon. Rich deep-blue lighting. INTRO Scene 2a: Huge kinetic word “Now” filling most of the frame. White typography fading into violet. A glowing purple horizon/platform beneath the word. Scene 2b: Large word “Distributed” across the upper part of the frame. Beneath it, a cinematic laptop displaying a premium dark dashboard, sitting in a moody purple-blue environment. Scene 2c: Centered laptop with the same dashboard. Several circular platform/account icons floating around it, connected visually as a distributed advertising ecosystem. PRIVATE MANAGEMENT INTRODUCTION Scene 3a: Minimal black frame with centered white word “Introducing”. A subtle warm gold glow beneath it. Scene 3b: Huge word “Private”, left aligned, partially fading into a dark champagne-gold silhouette / atmospheric background. Scene 3c: Centered premium logo lockup: “AdRevival” “Private Management” white logo on a black-to-gold gradient background. PRODUCT / USE CASE Scene 4a: Show a realistic premium dark-mode SaaS advertising management dashboard. Black interface with violet and gold highlights, cards, account data, navigation sidebar and analytical information. Scene 4b: Minimal typographic card emphasizing “6% of cashback”, with “6%” in champagne gold and the rest in white. Include a thin gold rounded graphical accent around the percentage. Scene 4c: The statement “6% of cashback on all platforms” centered horizontally. Surround it with several floating metallic gold coins / platform symbols in depth. Scene 5a: Another close-up of the AdRevival dashboard showing multiple managed advertising accounts or platforms. Dark interface, gold borders and violet CTA elements. Scene 5b: Blur the dashboard slightly in the background and place a centered premium confirmation/modal card in front. Add an oversized elegant gold cursor arrow pointing at the primary action button. Scene 6a: Minimal title card: “The highest level” with “highest level” subtly emphasized in warm gold. Scene 6b: A refined graphical interpretation of: “The highest level of management in every platform” using two small circular gold platform/account icons above or beside the typography. SUPPORT Scene 7a: A floating dark chat/message bubble outlined in gold saying: “Can I manage multiple platforms?” against a luxurious black-to-gold background. Scene 7b: Split editorial layout. Large bold typography: “24/7 Support” on the left, and a realistic dark chat-support interface screenshot on the right. Scene 8a: Minimal typographic transition: “No More > Insiders Chat Support” where “No More” is highlighted in gold and the remaining words are muted white/gray. Scene 8b: Create a second variation of this frame with slightly different line breaks and hierarchy, still emphasizing “No More” in gold. Scene 9a: Minimal atmospheric frame with the phrase: “Keys to the castle” in muted champagne-gold typography on a black background with fine dot-grid texture. Scene 9b: Minimal title card: “Unbeatable Cashback” with subtle gold gradient typography and dark grid texture. CTA Scene 10a: Final premium brand end card. Centered AdRevival circular logo and wordmark: “AdRevival” “Private Management” on a black-to-champagne-gold cinematic background. Small line of copy underneath: “Only with AdRevival Private Management.” BOARD LAYOUT: Present all of these keyframes simultaneously on one large dark creative-direction board. Frames should be approximately the same size and aligned to an invisible grid. Use generous spacing between groups. Hook on the far left, Intro next, Use case occupying the largest center section, CTA on the far right. Some columns may contain 3–5 vertically stacked frames. Leave intentional empty space around the storyboard. Add a few tiny production notes above the central columns, like a real creative director’s working board. The overall image should look like an exported Figma storyboard / motion-design treatment made by a world-class branding studio. Extremely polished, coherent art direction, readable typography, consistent logo usage, accurate UI proportions, high fidelity, presentation-ready.

Sourcevia LaplaceYoung
Storefront glass: reflection plus etched type

Storefront glass: reflection plus etched type

An advertising photo of a flagship store's glass display window at dusk, showing exactly three identical handbags displayed in a row on a lit pedestal inside, with the brand name "NOIR & CO." etched in thin gold lettering on the glass itself, positioned so the etched text overlaps the reflection of the street lamps outside without losing legibility. The reflection on the glass must show a faint mirrored version of the street outside while the three handbags behind the glass remain sharply visible and unreflected-over, dusk color grading, editorial luxury retail photography. #GPtimage

by @opener_aiSource
Wooden chess knight 360° turntable

Wooden chess knight 360° turntable

"Create a photorealistic studio turntable video of a polished wooden chess knight. The knight has a minimalist carved horse silhouette: broad flat sides, a rounded elongated muzzle, a tiny dark eye, an angular ear, and a gently curved neck. A smooth, dark brown insert follows the mane along the back. The figure stands on a wide circular wooden base with several concentric stepped rings. Use warm brown wood with clearly visible vertical grain, softly rounded edges, and a glossy lacquer finish. Preserve the exact shape, proportions, wood grain, and dark mane insert throughout the video. The entire chess piece, including its base, rotates smoothly through one complete 360-degree turn around its vertical axis at a constant speed. It stays perfectly centered and firmly on the surface. The first and last frames match for a seamless loop. Keep the camera completely stationary, looking slightly downward at the piece. Show the entire object with a small margin above and below. Use a seamless light-gray studio background and floor, soft diffused lighting, gentle highlights on the lacquer, and a subtle contact shadow beneath the base. Duration: 3 seconds. Frame rate: 30 fps. Square 1:1 composition. No camera movement, zoom, cuts, wobbling, floating, deformation, changing proportions, sliding wood textures, flickering, additional objects, text, or logos."

by @higgsfieldSource

How to write great GPT Image 2.5 prompts

  1. 1Say what must not change. This is the one habit that separates 2.5 from every earlier model. "Keep the pose, lighting, camera angle and background exactly as they are; change only the jacket" gets you a changed jacket. Leave the lock out and the model re-imagines the whole frame.
  2. 2Change one variable at a time. Edits compound badly. Move the lamp, look, then relight — three prompts, three checkpoints. Asking for the move and the relight and a new colour grade in one go is how you lose the face.
  3. 3Put exact text in quotation marks. Typography is where 2.5 pulled ahead. Write the words you want rendered inside quotes, spell out the case, and name the type treatment: `the words "FIELD & FLOUR" in a bold condensed sans, centred`. Describing the text instead of quoting it gets you approximate letters.
  4. 4Assign every reference a job. With more than one input, say which is which: image 1 is the person, image 2 is the garment, image 3 is the location. Unassigned references get blended, and the blend is never the one you wanted.
  5. 5Photoreal comes from camera language. Lens length, distance, light source, time of day, film stock, depth of field. "50mm at eye level, soft coastal daylight, shallow depth of field, fine grain" reads as a photograph; "photorealistic, 8k, ultra detailed" reads as a render.
  6. 6For layouts, describe structure before content. Infographics, slides, UI and posters land far better when the prompt gives the grid first — how many panels, what sits where, what the margins do — and only then fills in the words and images.
  7. 7Write negatives as concrete objects. "No watermark, no extra fingers, no duplicated logo, no orange colour cast" beats "avoid AI artifacts". Name the artefact you keep getting and it usually goes away.
  8. 8Pick the tier on purpose. On apimodels.app every quality tier is priced differently (1K runs $0.01 to $0.08, an 8x spread), so the tier is both a detail and a cost decision. Low is for layout checks and thumbnails — fastest and most affordable. Medium is the sane default for most finished work. High is for skin, fabric, dense type and anything you will print. Ask for 4K only when the detail is really there to resolve — an upscale of a 1K idea looks like an upscale, and above the low tier 4K costs more than 2K.

Prompt formula

What must not change + subject + scene, light and lens + exact text in quotes + material and surface detail + layout or aspect ratio + concrete negatives

Example: Keep the model, her pose, the studio lighting and the seamless grey backdrop exactly as they are. Replace only the tote she is holding with a natural-canvas one, and print the words "FIELD & FLOUR" across it in a bold condensed sans, dark brown, centred and following the fold of the fabric. Match the existing shadow direction and the slight warmth of the key light. Keep the canvas weave visible. No new props, no logo anywhere else, no change to her hands. 4:5.

Frequently asked questions

What is GPT Image 2.5, and what is the difference between Flare and Sunburst?

GPT Image 2.5 is the image model OpenAI released on 9 September 2026. The API exposes two variants. Flare holds GPT Image 2 quality and editing ability while cutting latency by roughly half — it is the one to reach for by default. Sunburst spends longer per image and returns more precision, which shows up in skin, fabric, dense typography and complicated product surfaces. Both can return a transparent background and generate natively at 1K, 2K and 4K rather than upscaling a smaller render. Neither takes a mask — you scope an edit by naming in the prompt what must change and what must not.

Are these prompts free to use?

Yes. Every prompt on this page is free to copy and run wherever you have access to GPT Image 2.5 — no login, no payment, no attribution required by us.

Where did these prompts come from?

Three places, and each card says which. Some are the worked examples from OpenAI's own image prompting guide. Some were collected into public GitHub repositories, and we credit both the prompt author and the person who did the collecting. The rest are from creators' own posts on X, credited and linked back. Copyright stays with whoever wrote the prompt; we only curated and paired each one with the image its source published. If you wrote one of these and want it removed or the credit corrected, email us.

Are the images on this page generated by GPT Image 2.5?

They are the images the source published alongside the prompt, not images we re-generated. Where a source only showed a design target or a side-by-side against an older model, the card says so rather than passing it off as that prompt's output. Eight prompts have no image because their source never published one, and we left those blank instead of substituting something else.

Where can I run GPT Image 2.5?

On apimodels.app, as gpt-image-2.5-flare and gpt-image-2.5-sunburst, live since 9 September 2026. Pricing runs on two axes, resolution and quality — $0.01–$0.08 at 1K and $0.020–$0.15 at 2K on both models, and at 4K $0.02–$0.3 on Flare against $0.02–$0.47 on Sunburst (low to max) — so the tier you pick is both a detail and a cost decision. Charged only on success, with the same key covering GPT Image 2, Nano Banana, Seedream, Seedance, Kling, VEO and 140+ other models. Leave the quality parameter off and Flare defaults to medium, Sunburst to high.

Do these prompts still work on GPT Image 2 or other image models?

Mostly, with one caveat. Composition, lighting, typography and style instructions transfer fine — that part of the craft is model-agnostic. What does not transfer is the strict edit locking: telling an older model to change only the jacket and leave everything else untouched gets you a partly re-imagined frame. If you are running these on something other than 2.5, expect to iterate more on the editing prompts and about the same on the text-to-image ones.

Run these prompts on apimodels.app

GPT Image 2.5 is live here alongside 160+ other image, video, audio and language models — one API, one key, pay as you go, with $0.10 of trial credit on sign-up.

This library is open source on GitHub

The whole library is downloadable as JSON and CSV, one page per category with the full prompt text. Every entry credits the person who wrote it and links to the source; where a prompt reached us through someone else's collection, the collector is credited too. Prompt text belongs to its authors — the MIT licence covers the curation only. Open an issue to add a prompt or correct a credit.

View on GitHub

Other prompt libraries