Skip to main content

Figurines

Chibi Doll / Clay Figure

A chibi doll of anyone: large blank head, tiny body, matte vinyl or clay texture, painted eyes and a simple wooden desk scene.

from 3 credits4 prompt setsNeeds one photoNano Banana 2Nano Banana 2 Lite

model: grok-imagine-image-2 → nano-banana-2-lite. The set runs image-to-image from a pet photo; grok-imagine-image-2 is routed text-to-image only on this site (apps/api/src/media/kie.ts), so the reference would be dropped. nano-banana-2-lite runs image-to-image at 3 credits and accepts the same 4:5 ratio.

Example results

Reference image

Fictional adult woman with coiled hair against a plain studio background

Generated result

A generated example for the Chibi Doll / Clay Figure image prompt template.

Chibi Doll / Clay Figure

GPT Image 2.5 Flare · 1K · 1:1 · 6 credits

Reference images were generated for this demonstration; no real person or customer photo was used.

Generation prompt
Turn the person in the reference image into a chibi collectible doll standing on a wooden desk. Carry over the silhouette of their haircut, the hair colour and the dominant colour of what they are wearing, and compress everything else into collectible toy language. Proportions: head roughly one third of the total height, torso and legs together about two head heights, arms and legs short and slightly tapered, hands shaped like simple mittens. Face: two big painted oval eyes each carrying one highlight dot, no lashes, one thin painted brow line, a tiny painted mouth, flat matte cheeks with one soft blush patch. Material: matte polymer clay with faint tooling marks and a soft satin sheen, a few dust specks on the shoulders, a thin seam line where the head meets the neck. Hair: sculpted as three or four solid masses with a visible parting and one loose strand. Pose: standing with the weight on one leg, both hands lifted a little as if about to wave, head tilted a few degrees. Scene: a wooden desk surface with a small cup and a closed notebook, both softly out of focus behind the doll. Lighting: soft window light from the left with a gentle fill, one soft shadow to the right of the doll. Photographic macro product shot, shallow depth of field, no illustration style, no HDR.

What it is used for

  • Follow the chibi doll trend on social media, where dedicated tutorials exist for French, Spanish and Russian audiences.
  • As a gift: turn someone into a desk ornament for them to keep.
  • Content material for craft and clay accounts.
  • Turn a couple or two friends into a matched pair of figures from the same series.

Prompt sets

Each set is a complete prompt. Copy it, or open it in the studio with the model, aspect ratio and resolution it was written for.

Desk Companion Figure

The standard setup: one figure on a desk scene, the most reliable option.

5 credits

Nano Banana 2image → image1:1 · 2K

Reference photos: One front-facing portrait, a half-body shot is enough. What the figure needs to keep is the hairstyle silhouette and the main clothing color, not skin detail.

Turn the person in the reference image into a chibi collectible doll standing on a wooden desk. Carry over the silhouette of their haircut, the hair colour and the dominant colour of what they are wearing, and compress everything else into collectible toy language. Proportions: head roughly one third of the total height, torso and legs together about two head heights, arms and legs short and slightly tapered, hands shaped like simple mittens. Face: two big painted oval eyes each carrying one highlight dot, no lashes, one thin painted brow line, a tiny painted mouth, flat matte cheeks with one soft blush patch. Material: matte polymer clay with faint tooling marks and a soft satin sheen, a few dust specks on the shoulders, a thin seam line where the head meets the neck. Hair: sculpted as three or four solid masses with a visible parting and one loose strand. Pose: standing with the weight on one leg, both hands lifted a little as if about to wave, head tilted a few degrees. Scene: a wooden desk surface with a small cup and a closed notebook, both softly out of focus behind the doll. Lighting: soft window light from the left with a gentle fill, one soft shadow to the right of the doll. Photographic macro product shot, shallow depth of field, no illustration style, no HDR.
Negative prompt
realistic human proportions, uncanny face, glossy plastic, tangled hair strands, cluttered background, extra limbs, text, watermarks

Reference only — the image models available here do not take a separate negative prompt, so this list is guidance for editing the prompt above.

Why this works: Main variant. Generate this one first to confirm the head-to-body ratio, then move on to the pair or the pet variant; once the proportions are set, the other variants can reuse the same proportion sentence.

Use this prompt

Matched Pair Figure

For a couple, two friends or two teammates made as two figures in the same style.

5 credits

Nano Banana 2image → image3:4 · 2K

Reference photos: Two portraits, one of person A and one of person B. The roles are assigned by first and second image, so do not upload them in the wrong order.

Create two matching chibi collectible dolls standing side by side on a light wooden surface, using the first reference image for the doll on the left and the second reference image for the doll on the right. Both dolls share exactly the same proportions and material, which is what makes them a set: the head is about one third of the total height, the body is about two heads tall, short tapered limbs with mitten hands. Faces follow the same simplified template: two large painted oval eyes with one highlight dot each, a single thin eyebrow line, a tiny painted mouth, one soft blush patch on each cheek; keep each person's hair colour and hairstyle silhouette distinct so the two dolls are identifiable. Material: matte polymer clay with faint tooling marks, soft satin sheen, a thin seam line at the neck. Clothing: simplified rounded garments in each person's main colour with one contrasting accessory each, the left doll holding a small book, the right doll holding a small cup. Pose: both turned slightly toward each other, the left one leaning a few degrees. Scene: a light wooden surface, a plain cream wall behind, one small plant out of focus at the right edge. Lighting: soft diffused light from the upper left, gentle contact shadows, no harsh reflections. Vertical 3:4 framing with both dolls fully visible and the same scale. Photographic macro shot, no illustration style, no text.
Negative prompt
mixed proportions between the two dolls, three dolls, realistic faces, glossy plastic, cluttered background, extra hands, text, logos

Reference only — the image models available here do not take a separate negative prompt, so this list is guidance for editing the prompt above.

Why this works: Pair variant. The two dolls must share the same three passages for material, proportions and face template, with only the hairstyle and the colors changed; otherwise they come out in two different styles.

Use this prompt

Pet Chibi Figure

Make a pet figurine with the same technique.

3 credits

Nano Banana 2 Liteimage → image4:5 · 1K

Reference photos: One front-facing photo of a pet; a cat or a dog both work. The fur color and markings are kept, and the body is simplified throughout.

Turn the pet in the reference image into a chibi clay figurine standing on a desk. Keep the fur colour, the markings and the overall head shape recognisable; simplify everything else into toy form. Proportions: the head is about one half of the total height, the body is short and rounded, the legs are four small tapered stubs, the tail is a single curved mass. Face: two large painted oval eyes with one highlight dot each, a small painted triangular nose, a simple painted smile line, whiskers made of four short sculpted lines on each side, one soft blush patch on each cheek. Material: matte polymer clay with faint tooling marks and a soft satin sheen, one thin seam line along the back, tiny dust specks on the paws. Ears: sculpted as two solid masses, one slightly bent, with a lighter inner surface. Pose: sitting upright with the front paws together, head tilted, as if waiting for a treat. Scene: a desk corner with a small ceramic bowl and a folded cloth, both softly out of focus. Lighting: soft window light from the left, one soft shadow to the right, a gentle highlight on the nose. Macro product photography, shallow depth of field, no illustration style, no text.
Negative prompt
realistic fur rendering, uncanny animal face, glossy plastic, messy background, extra legs, text, watermarks

Reference only — the image models available here do not take a separate negative prompt, so this list is guidance for editing the prompt above.

Why this works: Pet variant. Describe the fur as sculpted masses ('solid masses') rather than in fur detail, or the image falls into realistic rendering.

Use this prompt

Text-Only Chibi Sample

For page samples and low-cost stress testing.

3 credits

Nano Banana 2 Litetext → image1:1 · 1K

A macro product photograph of a generic chibi clay figurine standing on a wooden desk. Proportions: the head is about one third of the total height, the body is about two heads tall, arms and legs are short tapered stubs with mitten hands. Face: two large painted oval eyes each with one highlight dot, a single thin painted eyebrow line, a tiny painted mouth, one soft blush patch on each cheek, no eyelashes. Hair: sculpted as three solid masses with a visible parting and one loose strand. Material: matte polymer clay with faint tooling marks, a soft satin sheen, a thin neck seam, a few dust specks. Clothing: a simplified rounded jumper in dusty blue with a small contrasting collar. Pose: standing with the weight on one leg, one hand raised slightly. Scene: a wooden desk surface, a small cup and a closed notebook out of focus behind the doll, a plain cream wall as the background. Lighting: soft window light from the left with a gentle fill, one soft shadow to the right. Shallow depth of field, photographic realism, no illustration style, no text, no logos.

Why this works: Text-only sample variant, the cheapest option. It is a fallback for pages that have no source material, and its proportion and face-template sentences match the main variant so the two can be compared side by side.

Use this prompt

If the result is wrong, change this first

The bigger the figure gets, the more human it looks, and the head-to-body ratio drifts.

Write the proportions as numbers: 'the head is about one third of the total height and the body is about two heads tall'; a proportion sentence works better than any adjective.

The face lands in the uncanny valley, either too realistic or with misplaced features.

Simplify explicitly: 'eyes as two simple painted ovals with a single highlight dot, no eyelashes, tiny painted mouth, no eyebrows or a single thin line'.

The surface looks like a plastic toy: too shiny and too fake.

Swap in material words: 'matte polymer clay with faint fingerprint tooling marks', 'a soft satin sheen', 'tiny dust specks'; delete 'glossy' and 'plastic'.

The hair is one shapeless mass.

Write 'sculpted hair in three or four solid masses with a visible parting and one loose strand'; do not state a number of hair strands.

The background is too busy, like a figurine photo studio.

Limit the background to two props, such as a glass of water and a closed book, and write 'shallow depth of field with the background softly out of focus'.

Notes

  • Chibi doll failures almost always come down to the head-to-body ratio and the eye description. Writing those two as numbers and shapes works far better than piling up adjectives.
  • The cheapest approach for this template is to iterate on the proportions with a cheap model (nano-banana-2-lite / grok-imagine-image-2) and move to nano-banana-2 once the final look is set.