Start free

AI Art Generator

Describe what you want to see and name the medium: watercolor, oil, ink, linocut, stained glass, pixel art, flat vector or anime. Every piece on this page came out of Leap, and each one opens in the studio with its prompt.

Model

Start free, no card needed. You sign in first; your prompt waits for you in the studio.

Real pieces, with their prompts

Pick a style, then make it yours

Each piece came back as you see it, from the prompt under it. Press Make this to open the same model with the prompt in the box, then swap the subject and keep the style.

Watercolor, oil and gouache

Each prompt names the paint and the surface first, then the scene, then what the paint does up close: washes bleeding into each other and bare white paper, visible brushwork and cracked varnish, matte color with dry-brush texture, smooth glazes and long shadows. The still life puts a gas station hot dog in an old Dutch master's light.

  • Watercolor: the fishmonger and the gull, made on Leap

    Watercolor: the fishmonger and the gull

    Nano Banana Pro

    A loose watercolor painting on rough cold-press paper: a fishmonger in her sixties in yellow rubber boots hoses down the market floor at dawn while a gull steals a sardine off the ice behind her back. Wet-in-wet washes bleeding into each other, granulating blue shadows, bare white paper left for the light, faint pencil lines showing through. No text.

    Make this

    $0.1907 for a 1K image on Nano Banana Pro

  • Oil: still life with a gas station hot dog, made on Leap

    Oil: still life with a gas station hot dog

    FLUX 3 Image

    A Dutch Golden Age still life oil painting: a half-eaten gas station hot dog in its paper tray, a crushed energy drink can with no label, a vape pen, a peeled lemon with its rind spiralling off the table edge, and a melting ice cream sandwich on a pewter plate, one fly on the rim. Dark background, window light from the left, visible brushwork, fine cracks in old varnish.

    Make this

    $0.0264 for a 1K image on FLUX 3 Image

  • Gouache: grandma, the goat and the sidecar, made on Leap

    Gouache: grandma, the goat and the sidecar

    Seedream 4.5

    A gouache illustration from a 1960s picture book: a grandmother in a leather jacket rides a motorcycle along a road through a sunflower field, a goat in the sidecar wearing round goggles. Matte opaque paint, simple flat shapes with dry-brush texture, a muted palette of mustard, teal, brick red and cream. No text.

    Make this

    $0.044 for an image on Seedream 4.5

  • Surrealist oil: the bus stop in the desert, made on Leap

    Surrealist oil: the bus stop in the desert

    FLUX 3 Image

    A surrealist oil painting: a man in a grey business suit waits at a bus stop in an empty desert, holding an umbrella that rains only on himself, goldfish swimming out of his open briefcase into the puddle at his feet. Smooth glazed paint, long hard shadows, a pale turquoise sky with one cloud.

    Make this

    $0.0264 for a 1K image on FLUX 3 Image

Ink, charcoal and pen

A drawing prompt names the tool and the paper (ink on rice paper, charcoal on toned grey paper, blue ballpoint in a sketchbook) and the marks it leaves: one wet sweep of the brush, smudges and eraser marks, cross-hatching and a coffee ring.

  • Ink: a sumo wrestler on the last train, made on Leap

    Ink: a sumo wrestler on the last train

    GPT Image 2

    A sumi-e ink painting on rice paper: a sumo wrestler asleep upright on a subway bench late at night, his belly drawn in one wet sweep of the brush, the train only a few dry strokes behind him. Black ink and soft grey washes, vast empty paper around him. No text, no seal.

    Make this

    $0.0563 for a 2:3 image on GPT Image 2

  • Charcoal: between rounds, made on Leap

    Charcoal: between rounds

    FLUX 3 Image

    A charcoal drawing on toned grey paper: a middle-aged boxer slumped on the stool between rounds, his cornerman pressing a bag of ice to his swollen eye. Smudged blacks, white chalk highlights, loose gestural lines, fingerprints and eraser marks left in the charcoal.

    Make this

    $0.0264 for a 1K image on FLUX 3 Image

  • Ballpoint: the night bus sketchbook, made on Leap

    Ballpoint: the night bus sketchbook

    GPT Image 2

    A blue ballpoint pen drawing across a sketchbook spread: quick gesture studies of passengers asleep on a night bus, a teenager in headphones, a nurse in scrubs, an old man holding a cake box on his lap. Cross-hatched shadows, a few loose construction lines, a coffee ring stain on the paper. No writing.

    Make this

    $0.0563 for a 3:2 image on GPT Image 2

Prints, glass and collage

Say how the piece was made and what that leaves behind: carved gouge marks and uneven ink for a linocut, thick lead lines and sun through colored glass for a window, scissor edges and halftone dots for a collage. Each ends by ruling out lettering, so nothing gets written in.

  • Linocut: the fox and the food truck, made on Leap

    Linocut: the fox and the food truck

    Grok Imagine 2.0

    A two-color linocut print in black and vermilion ink on cream paper: a fox raiding a tipped-over trash can behind a food truck at night, a one-eyed cat watching from the truck roof. Rough carved gouge marks, chunky shapes, uneven ink with paper showing through, the slight wobble of hand printing. No lettering.

    Make this

    $0.066 for a 1K image on Grok Imagine 2.0

  • Stained glass: the delivery rider, made on Leap

    Stained glass: the delivery rider

    GPT Image 2.5 Sunburst

    A stained glass window: a food delivery rider on an e-bike in the rain, a big insulated backpack on his back, a halo of streetlight around his helmet like a saint. Thick black lead lines, jewel-toned glass in ruby, cobalt, emerald and amber, afternoon sun shining through. No text.

    Make this

    $0.0858 for an image on GPT Image 2.5 Sunburst

  • Collage: the conch megaphone, made on Leap

    Collage: the conch megaphone

    Nano Banana Pro

    A paper collage made from torn vintage magazine pages: a woman shouting through a megaphone made from a conch shell, her hair a mess of torn orange paper, waves cut from blue paper behind her. Visible scissor edges, slight paper shadows, halftone dots in the cutouts. No readable text.

    Make this

    $0.1907 for a 1K image on Nano Banana Pro

Pixel art, vector and anime

Digital styles need their rules spelled out: a limited palette and crisp square pixels with no blur, flat shapes with no gradients and a short list of colors, or the decade of the anime and the look of its film.

  • Pixel art: tacos under the overpass, made on Leap

    Pixel art: tacos under the overpass

    GPT Image 2.5 Flare

    16-bit pixel art with a limited palette of about 24 colors, crisp square pixels and no blur: a late-night taco stand under a highway overpass, the cook flipping tortillas on a steaming griddle, three customers in hoodies eating standing up, a stray dog waiting for a dropped bite, string lights overhead. Side-on view like a game level. No letters or numbers.

    Make this

    $0.0858 for an image on GPT Image 2.5 Flare

  • Flat vector: gridlock from above, made on Leap

    Flat vector: gridlock from above

    Recraft V4.1

    A flat vector illustration seen from straight above: a bike courier weaving through a gridlocked intersection of yellow taxis and a city bus, long afternoon shadows, pigeons scattering. Bold flat shapes, no gradients or texture, crisp geometric edges, a palette of tomato red, teal, mustard, cream and navy. No text.

    Make this

    $0.0385 for an image on Recraft V4.1

  • 90s anime: the raccoon at 3 a.m., made on Leap

    90s anime: the raccoon at 3 a.m.

    Grok Imagine 2.0

    A frame from a 1990s cel-animated anime film: at 3 a.m. a tired convenience store clerk with bleached hair and dark circles under her eyes stares down a raccoon standing on the counter, clutching a stolen rice ball. Hand-painted background, harsh fluorescent light, slightly faded colors and film grain. No readable text anywhere.

    Make this

    $0.066 for a 1K image on Grok Imagine 2.0

Pick the model for the medium

Each model has its own hand. One account runs them all, so you can try the same prompt on two or three and keep the one you like.

What it costs

Start free, then pay per image

A new account comes with $2 of credit, and no card is needed. After that you add credit when you want more and pay for each image you make. There is no subscription, and an image that fails costs nothing.

ModelFromWhat $2 makes
Nano Banana Pro$0.19About 10 images
FLUX 3 Image$0.026About 75 images
GPT Image 2$0.056About 35 images
GPT Image 2.5 Flare$0.07About 28 images
Grok Imagine 2.0$0.044About 45 images
Recraft V4.1$0.039About 51 images
Seedream 5.0 Pro$0.039About 51 images

Each price is one run at the model's lowest-priced settings, from the live catalog. Larger sizes, longer clips and sound cost more, and the studio shows the exact price before every run. How pricing works.

From a photo, or for a job

More ways to make art

Recipes do the prompting for you: turn your own photo into a painting, or make art for a tattoo, a coloring book or a record.

Photo to drawing
Your photo, drawn or painted by hand: pencil, charcoal, ink, watercolor or oil.
Photo to cartoon
Turn a photo of a person or pet into an illustrated character that still looks like them.
Anime
Describe a character or a scene and get it drawn as anime, or add a photo to be in it.
Royal pet portrait
Your pet as the subject of a regal oil painting, in a crown, a uniform or a lace collar.
Tattoo design
Describe the tattoo you want and get the design, ready to take to your artist.
Coloring page
Black and white line art of anything you describe, ready to print and color.
Album cover
Square cover art for an album or single, with the artist and title set in the design.
AI image generator
For photos, film stills, posters and product shots rather than paintings.

Questions

Questions people ask

Is it free?
You can start free: a new account comes with $2 of credit, and no card is needed. That is about 75 pieces with FLUX 3 Image or 10 with Nano Banana Pro. After that you add credit when you want more and pay per image. There is no subscription, and the welcome credit is one per person.
Which model should I use for which style?
Each piece on this page names the model that made it, so start from the one closest to what you want. Here Nano Banana Pro painted the watercolor, the impasto diner and the collage, and FLUX 3 Image the still life, the surrealist desert and the charcoal, for a few cents each. GPT Image 2 drew the ink, the ballpoint sketchbook and the woodblock print, GPT Image 2.5 Flare the pixel art, Grok Imagine 2.0 the linocut and the 90s anime, and Recraft V4.1 the flat vector. You can run the same prompt on two or three and keep the best.
How do I get a particular style?
Name the medium first, then the scene, then how the medium looks up close: the paper, the brush marks, the ink, the pixels. Press Make this under any piece to open its prompt in the studio, then change the subject and keep the rest.
Can I make art in a famous artist's style?
Describe the look instead of naming the artist: the medium, the palette, the brushwork and the era. That gets closer to what you want, and it keeps clear of other people's rights. Leap's Acceptable Use Policy rules out prompting with copyrighted material you don't have the rights to use.
Can I sell what I make or print it?
Leap's Terms say that, as between you and Leap, you own the results the studio makes for you, to the extent the law allows, and they set no personal-use-only limit. In some countries AI-made images may not be protected by copyright, so nobody can promise a result is unique. For a large print, pick a model and size made for it: Nano Banana Pro goes up to 4K, and Recraft V4.1 Pro makes four-megapixel images and vectors.
Can I turn my photo into a painting?
Yes. Photo to drawing turns a photo into pencil, charcoal, ink, line art, colored pencil, watercolor or oil and keeps the people in it. Photo to cartoon redraws it as a cartoon, and Anime makes you the main character of a scene.
Do you train AI on my prompts or images?
No. Leap does not use your prompts, uploads or results to train AI models, and does not use them in its marketing without asking you first.
What happens if an image fails?
It costs nothing. Only a run that succeeds is charged, and you see the price on the Generate button before you press it.

Your first piece, from one line.

Start free, no card needed. Every image is priced before you press Generate.

Make art