Upload a photo · Keep the scene · One pass
Drop in a photo and it is repainted, not filtered: the sky becomes washed gouache with big soft clouds, the buildings pick up visible brushwork, faces are redrawn as clean line art, and the whole frame is re-lit as though it had been painted for an animated film. Your first render is on us.
Both frames come from our own run on this page's model while building it. Drag the divider: the green linen shirt, the floral dress, the canvas tote and the café awning behind them all survive, and the pavement, the sky and the passers-by have been repainted rather than tinted. Every output is AI-generated, so your result varies with your photo and your prompt — read this as what the tool can hold, not a fixed outcome.
Drag to compare before and after
It is the visual grammar of hand-drawn Japanese animation: watercolour and gouache backgrounds laid down in visible strokes, skies that carry most of the colour in a frame, flat cel shading on characters with one soft shadow shape, and faces reduced to clean confident lines with large clear eyes.
The thing that separates a good repaint from a bad one is not how painterly the sky looks — it is whether the picture is still yours afterwards. A filter dials up saturation and softens edges and leaves a photograph underneath. A repaint regenerates every surface in the style, which means it can also quietly replace your face, your jacket or the building behind you. This page is built around keeping those.
Creative engine
Upload the photo, keep the preset prompt or name the colours you want held, pick the shape. The cost estimate shows before you generate and a failed run is refunded automatically unless it broke the content policy.
The Ghibli style lives in the sky and the scenery far more than in the characters. A photo with something behind it — a street, a hillside, an open horizon — has room for the brushwork; a tight headshot against a blank wall has almost none.
Green shirt, floral dress, denim jacket. Two or three colour nouns are usually enough to anchor the repaint, and without them the model is free to redress everyone.
The output is an original picture in a widely used animation idiom. It is not affiliated with, endorsed by or produced by any film studio, and it does not generate anybody's copyrighted characters.
For the same photo in other painted registers, oil painting portrait works in thick visible impasto, photo to claymation rebuilds it in modelling clay, and the pixar filter renders it in 3D instead of painting it flat.
The same prompt on a full-length photo shot against nothing in particular, to see where the style puts its effort when the scene hands it a blank.
The source was a woman standing against a plain studio backdrop. Given no scenery to repaint, the model built one — watercolour hills, a peach and blue washed sky, brushwork left visible in the clouds — while the denim jacket, the pale jeans, the white trainers and the stance came through unchanged. Useful to know in both directions: a blank background will be filled with something invented, so if the real one matters, say so in the prompt.
Open the image studioThree things worth stating outright. Leave them vague and you get a lovely painting of somebody else's afternoon.
A street, a garden, a hillside, a window. The idiom carries most of its charm in the scenery, and a picture with depth behind the subject gives the repaint somewhere to put the brushwork instead of inventing it.
Ask for painted washes, visible brushwork and flat cel shading, and say explicitly that photographic texture should go. Left open, the model often lands on a soft, glowing photo that reads as a preset rather than a painting.
The colour of a shirt, the length of someone's hair, the sign on the shopfront. Named, they anchor the repaint. Unnamed, they are fair game, and that is where the picture stops being yours.
Three places a painted version of a photograph earns its keep.
A holiday photo or a family shot repainted holds up at print size in a way a filtered version does not, because the detail is drawn rather than smeared.
The look reads clearly at small sizes — flat shapes, strong line art, high-contrast sky — which is the same reason the emote maker works at icon scale.
One painted frame from the day, alongside the photographs, is a common request; the wedding photo generator covers the same occasion in a photographic register.
What it keeps, what it will not do, and what a repaint costs.
No. This is an independent tool that repaints photographs in a widely used hand-drawn animation idiom. It is not affiliated with, endorsed by or connected to any film studio, and it does not generate studio-owned characters or scenes from any film.
Usually yes, and it depends more on your photo and prompt than on the model. Faces are simplified into line art by design, so what carries the recognition is everything around them — hair colour and length, clothing colours, poses, the setting. Name those in the prompt and they hold. Both examples on this page kept every one of them.
Because the prompt left the medium open. Add painted gouache washes, visible brushwork, flat cel shading and no photographic texture, and it moves back to a painting. A glowing, slightly smeared photograph is what happens when the model is asked to stylise rather than to repaint.
It does better with one. The idiom carries most of its charm in backgrounds, so a street, a garden or an open horizon gives the repaint somewhere to put its brushwork. A tight headshot against a blank wall leaves it nothing to paint, and as our second example shows, it will invent scenery rather than leave the space empty.
Not on this page, which produces a single still. For a moving version the usual route is to repaint one frame here and then hand that image to a video model — the sketch to video page walks through the same handover for line art.
The first ones are covered by the free credits on the account, and anonymous visitors get a run before signing up at all. After that each repaint draws from your balance like any image, the estimate for your model and size appears in the generator before you start, and a failed run is refunded automatically unless it broke the content policy.
One credit pool covers Nano Banana images and Seedance and Veo video. The cost shows before every run, and failed jobs are not charged. Use a subscription for ongoing work, or a one-time pack when you just need to top up.
What's included
Secure checkout by Stripe. Card details never touch our servers.
What's included
Secure checkout by Stripe. Card details never touch our servers.
What's included
Secure checkout by Stripe. Card details never touch our servers.
What's included
Secure checkout by Stripe. Card details never touch our servers.
One-time top-ups — buy extra credits any time you run low.
What's included
Secure checkout by Stripe. Card details never touch our servers.
What's included
Secure checkout by Stripe. Card details never touch our servers.
What's included
Secure checkout by Stripe. Card details never touch our servers.
Charges appear as “KAVEL AI” on your card statement. You can cancel any time from Settings → Billing; cancellation takes effect at the end of the period you already paid for.
Powered by
Use one balance across every supported image and video model — eleven of them today, listed below — and check the credit cost before each request.
Video models
Image models
Model credit guide
You're only charged for successful generations. The exact estimate in the generator varies by model, length, resolution, audio, and number of images.
| Type | Model | Credit cost |
|---|---|---|
| Video | Seedance 2.5 | The long-take tier, priced per second: 5s 480p ≈ 211 credits, 5s 720p ≈ 473. A full 30s take runs ≈ 1,260 at 480p and ≈ 2,835 at 720p. Supplying a reference clip lowers the per-second rate. |
| Video | Seedance 2.0 | 5s 720p image-to-video ≈ 188 credits; text-to-video ≈ 308 credits. Scales with resolution and length. |
| Video | Seedance 2 Fast | Faster and lower cost. 5s 720p text-to-video ≈ 248 credits. |
| Video | Seedance 2 Mini | The cheapest Seedance tier. 5s 720p ≈ 154 credits, 5s 480p ≈ 72 credits. |
| Video | Seedance 1.5 Pro | Audio doubles the rate. 5s 720p ≈ 27 credits silent, ≈ 53 with audio; 1080p ≈ 57 and ≈ 113. |
| Video | Veo 3.1 | Billed per video, not per second (Lite tier). About 45 credits at 720p and 53 at 1080p. |
| Video | Kling 3.0 | Audio raises the rate. 5s 720p ≈ 105 credits silent, ≈ 150 with audio; 1080p ≈ 135 and ≈ 203. |
| Video | MiniMax H3 | Fixed 2K, no resolution ladder. Priced per second — a 5s clip ≈ 158 credits. |
| Image | Nano Banana 2 | Generate or edit from text and images. 20 credits per 1K image, 30 at 2K, 45 at 4K. |
| Image | Nano Banana Pro | Consistent run times across generations. 30 credits per 1K or 2K image, 50 at 4K. |
| Image | Nano Banana 2 Lite | Faster, simpler variant, and the everyday editing price: a flat 8 credits per image at every resolution. |
| Image | GPT Image 2 | The lowest-cost premium image model here. 10 credits per 1K image, 15 at 2K, 30 at 4K. |
| Image | Seedream 5.0 Lite | Flat pricing — 20 credits per image whether you generate or edit. |
Three tools that take the same picture and change what it is made of.
One picture with a background, and the two or three things in it that must not change. The repaint is one pass.
A real edit made with this tool — drag your own photo in on the left, the prompt is already written, then press Generate.
