22sang

목부터 ...

[Ciri @Ciri_ai]

GPT Image 2 on Chatgpt

Prompt: Please transform the entire image into a single Decorative Folk Flat Illustration with Doodle elements. Use a bold and playful color palette, completely different from the original image. Simplify all details into clean, flat shapes with a handmade, slightly imperfect feel, as if drawn on a sheet of white paper. The overall style should look cute, childlike, and whimsical

[Mahnoor Fatima @MahnoorAi12]

From photo to a warm handmade editorial collage ✨

Transformed this portrait into a minimal magazine cutout illustration with soft paper textures, torn edges, cozy scrapbook elements, and a clean Korean editorial aesthetic.

Made with GPT Image 2. On chatGpt
Prompt ⤵️

[Insomniak 🐦 @insomniak_77]

Margaret Qualley

[Sairah @Sairah_0]

GPT Image 2 On ChatGPT

Prompt:

Create a stylized gouache and marker illustration based on the following artistic style. Do not invent a new scene, transform the user’s reference photo using these exact visual characteristics:

•Style: Flat folk-art gouache illustration with bold opaque color and a matte poster-paint finish.
•Anatomy & Lines: Thick confident outlines, simplified blocky shapes, flat color fills with minimal shading, slightly rounded and softened proportions.
•Facial Features: Stylized heavy-lidded eyes with a graphic winged eyeliner flick, thin arched eyebrows, a long thin nose drawn with a single curved line, and small flat closed-mouth smiling lips. Clean, minimalist, doll-like facial structure.
•Complexion: Warm peachy skin tone, no freckles, with prominent solid rosy-pink circular blush on the cheeks (flat color, not stippled or textured).
•Composition: Close-up cropped portrait, subject resting cheek on hand with elbow raised, seated inside a car. Interior rendered in flat navy-blue with dashed white stitch details on the seats and door panel. Through the window, a simplified scene with a small red cartoon car, a smiling cartoon driver, a yellow sun, scalloped blue-outlined clouds, and scattered decorative leaves/hearts/flowers. Negative space above the figure filled with whimsical folk-art motifs: stars, sprigs, dots, arches, and a flower, on a cream background.
•Color Palette: Warm tones (peach skin, mustard yellow, coral pink, brown hair) contrasted against cool accents (navy blue, sky blue, purple, green), with high saturation throughout and no gradients.

[Ingi Erlingsson 🪄 @ingi_erlingsson]

SapienOS™

Made in @ComfyUI with GPT Image 2.0, Wan 2.2 I2V + vace, @thesystms HUD, Florence2, Sapiens2 + @suno

[Ciri @Ciri_ai]

GPT Image 2 on Chatgpt

Prompt: Create a stylized watercolor and gouache illustration based on the following artistic style. Do not invent a new scene, transform the user's reference photo using these exact visual characteristics:

- Style: Modern gouache folk art illustration with hand-painted watercolor textures and a matte finish.
- Anatomy & Lines: Distorted anatomy, expressive blocky shapes, fluid outlines, and visible soft brushstrokes.

- Facial Features: Stylized closed eyes, thin arched eyebrows, a long sharp nose, and simple flat lips. Distinct, clean, and minimalist facial structures.
- Complexion: Warm skin tones, heavy distinct freckles, and prominent rosy textured cheeks with a stippled or dry-brush effect.
- Composition: Organic flowing composition, interwoven layers, shallow depth of field, and a soft pastel background with vibrant botanical overlays.

- Color Palette: Warm earth tones contrasted against cool botanical accent colors.

[M @0detobeauty]

Pola Dzik photographed by Yan Wasiuchnik, June 2026.

[Marie Ruiz-Vidal @RuizVidal7]

Vittorio Giardino

[Marie Ruiz-Vidal @RuizVidal7]

Jane Fisher

[Marie Ruiz-Vidal @RuizVidal7]

© Nicholas Fols

[ComfyUI @ComfyUI]

Creator seungho__yeo ( IG ) has been architecting a workflow that doesn't just generate images - it builds an entire visual universe.

Environment Board → Fashion Board → Depth Extraction → Seedance Animation → Kinetic Edit

One workflow. One visual language.

A full New York sequence - Manhattan streets, yellow cabs, subway stations, rooftop editorials - animated and cut with crash zooms, hyperlapses, infinite zooms, and motion blur smears.

The insight: stop thinking in single images. Start designing asset collections.

The more insert shots, props, fashion details, and architecture you build upfront, the more powerful the final edit becomes.

[soli @solisolsoli]

Photo shoot from Dolce & Gabbana, 1987

[ComfyUI @ComfyUI]

Using Seedance 2.0 for video-to-video transformations opens up some interesting possibilities: change the setting, restyle the wardrobe, and shift the entire mood of a scene without reshooting.

This example reimagines a simple clip as a classic noir bar sequence. 🥃🎥

To try this workflow, link below 👇

[Ingi Erlingsson 🪄 @ingi_erlingsson]

it was real until it wasn't ✨

@thesystms TimeSlice + HUD with GPT Image 2.0 + Krea 2 + Grok Image in @ComfyUI and @AdobeAE

🎶 @suno

[Larus Canus @MrLarus]

🎾 Tennis-core K-pop dance feels so fresh!

Made with ChatGPT-Image2 + Seedance from a 16-panel storyboard

Sunny outdoor court, pink tennis skirt, white cropped polo

Side shuffles, hair flips, spins, jumps, low poses and a sweet idol ending!

Workflow + prompts below 👇

[Susana Lázaro @SusanLCoen]

Kaia Gerber

📸 Lachlan Bailey

[Skye Freeman @Skyebrows]

Little finicky this one, but we got there

[Kiki @Mayz1169]

I really love this kind of summer ocean vibe 🌊

made with GPT Image 2 + Seedance 2

workflow⬇️