The family in-joke
Everyone in the group chat rendered as characters from a film that does not exist. It survives being sent round far longer than a normal filter.
One portrait in · A puppet out · No sign-up
Tool marks in the forehead, a seam where the jaw would open, cloth that is actually cloth — a frame from a film that was never shot.
A real run on this page's model. Drag the divider and look at the shirt rather than the face: the plaid survives as woven cloth while the head becomes clay, and that mismatch of materials is exactly what your eye is reading as stop-motion. Results vary run to run, so treat this as the range rather than a guaranteed output.


Drag to compare before and after
It rebuilds a person as a stop-motion puppet: a modelled clay head with visible making marks, expressive oversized eyes, and real miniature clothing, photographed on a lit set rather than rendered on a background.
Nearly everyone who searches for this is chasing a memory rather than an aesthetic. Stop-motion is the animation people watched on television as children and can still feel the texture of, and seeing your own face in that world lands somewhere between funny and unexpectedly warm. That is a different motive from wanting a nice avatar, and it changes what a good result looks like: the goal is not flattering, it is *convincing as a physical thing somebody made*.
The tells are material contrasts. A stop-motion puppet is never one substance — clay face, fabric shirt, wire inside, sometimes glass or resin for the eyes — and it is the boundaries between those materials that make the brain accept it. The second tell is imperfection: the surface is slightly uneven, the paint is slightly inconsistent, and there is often a faint seam at the jaw because that is where the head was opened to animate the mouth. A perfectly smooth, perfectly symmetrical result is a 3D render wearing a clay palette.
It runs in the browser on the free image editor, image-to-image, with credits to start and no account needed for a first look. One honest limit: expression range. Real puppets have a small vocabulary of faces because each one has to be sculpted, so a subtle half-smile tends to come back as either a neutral face or a full open grin. If you want the grin, bring a photo where you are already grinning.
Creative engine
Upload a portrait, choose how exaggerated the sculpt should be, and check the estimate before generating.
Clothing comes back as miniature fabric with its own weave and folds, which is the detail that separates a puppet from a sculpture.
Warm directional key and a shallow plane of focus, so the scale reads as small even before you register what the material is.
If you want the smooth art-object version rather than a film character, that is the AI clay filter; for a rendered animated look try the Pixar filter, and for flat vector output the AI sticker generator.
Three steps, and the expression you bring decides the character.
Puppets have big, readable faces. A wide grin, raised brows or a proper scowl all translate; a neutral passport look comes back neutral and lifeless.
Name the thumbprints, the jaw seam and the fabric clothing explicitly. Those are the details that separate a stop-motion puppet from a smooth clay sculpture.
A larger head and shorter limbs read more like a puppet; realistic proportions read like a statue. Say which you want, check the estimate and run it.
Three reasons, and none of them is a flattering photo.
Everyone in the group chat rendered as characters from a film that does not exist. It survives being sent round far longer than a normal filter.
Seeing what a character design would look like as a physical build — head proportions, costume materials — before touching any clay.
For the person who grew up on stop-motion and would frame a picture of themselves living in it.
How the puppet look is built, and what it does with your face.
Usually because everything came back as one material. A puppet is clay *plus* cloth plus glass eyes, and it is the join between them that sells it. Ask explicitly for real fabric clothing and eyes with a glass highlight, and the result stops looking carved and starts looking built.
No — this produces a still frame that looks like it came from a stop-motion film. If you want the character to actually move, take the result and run it through a video tool; a clay puppet animates well because a viewer expects slightly jerky, weighty motion from one anyway.
Because puppet faces have a small expressive vocabulary — each expression is a sculpted state rather than a continuous range — and the model leans toward the readable ones. Subtle expressions tend to resolve to neutral or to a full grin. Bring the expression you want in the source photo.
The shape yes, the strands no. Hair becomes sculpted clay masses or a wool-like material with the same silhouette and colour, because that is how a physical puppet's hair is made. Very fine or very intricate styles simplify the most.
Yes, and pairs often work better than singles because the scene starts reading as a shot from a film rather than a portrait. Both faces need to be clear in the source; a second person half out of frame usually comes back rebuilt into someone else.
It stays, but it takes on the set treatment — shallower focus and warmer light, so it reads as a miniature build behind the character. If you want a specific set instead, describe it, and expect the character to change slightly when the model has two things to invent at once.
No. It runs on the free image editor in the browser with credits to start, so you can put one portrait through and judge the result before deciding anything else.
The estimate is shown before you generate and moves with the resolution you choose. Failed generations are refunded automatically unless they broke the content policy, so testing a second expression costs nothing when it fails.
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. About 8 credits per 1K image, 12 at 2K, 18 at 4K. |
| Image | Nano Banana Pro | Consistent run times across generations. About 12 credits per 1K or 2K image, 21 at 4K. |
| Image | Nano Banana 2 Lite | Faster, simpler variant. A flat 8 credits per image at every resolution. Start here to test. |
| Image | GPT Image 2 | The lowest-cost image model here. About 3 credits per 1K image, 6 at 2K, 12 at 4K. |
| Image | Seedream 5.0 Lite | Flat pricing — about 8 credits per image whether you generate or edit. |
Thumbprints in the clay, cloth that behaves like cloth, and a set light falling across a face that is still recognisably yours.