FlashBoards LearnPanel
  1. Learn
  2. Formats & Exhibition
  3. AI Cartoon

Formats & Exhibition · AI animation

AI Cartoon

External-tool reference examples from the source experiments, not a claim of current FlashBoards model availability or installed automation. Use your own supported tools and authorized media; settings, timings and comparisons describe the tested runs.

Also called: AI animated short, AI cartoon episode, AI anime short, AI claymation, storyboard to animation

A cartoon episode from one picture and one request: draw the story as a nine-frame storyboard with your character, give it to MiniMax H3 Max Turbo with a prompt full of acting and lines, and get 15 seconds back with the voices and the music. Below, thirteen episodes with Mara and her sidekick Pip in thirteen styles, every storyboard and prompt, and the way we got here.

Watch first: the whole guide in 1:53, hosted by Yaro
Video script

Hi, I'm Yaro. Today we make a whole cartoon episode from one picture and one request. No pipeline. Not even a tiny pipe.

Meet the cast: Mara, our recurring character, and Pip, a tiny fox-raccoon invented entirely in words. He has never forgiven us.

Step one: draw the whole story as one storyboard. Nine numbered frames: setup, problem, gag, payoff. GPT Image 2.5 draws it from Mara's sheet.

Step two: one request to MiniMax H3 Max Turbo. The board goes in as the first image. No last frame. No audio. Just vibes.

The prompt does the acting. It says the board never appears on screen, describes both voices in words, walks through nine shots with seconds, emotions and lines in quotes, and asks for sound generated with the picture.

Fifteen seconds later... no, three to six seconds later, you get fifteen seconds of cartoon: two voices, foley, music that fits the style. The vacuum swallows Pip. He comes out a dandelion.

One catch: the board flashes for one to seven frames at the very start. Snip. Trimmed. Nobody saw anything.

The experiment covered thirteen styles, from claymation to ukiyo-e, with one storyboard per episode.

We tried the hard way first: a generation per shot, voices and effects from ElevenLabs. Stiffer acting, effects shouting over the voices. Dropped. For cartoons. I'm still allowed to talk.

So: draw the story, then animate it once. Every storyboard, prompt and episode is right below. Scroll down. Pip is still mad about the vacuum.

What it does
A 15-second animated episode with your character, voiced and scored, in any style.
Use
GPT Image 2.5 for a 3×3 storyboard, MiniMax H3 Max Turbo image-to-video for the episode and its sound.
Generation time
3–6 s of generation per 15 s episode.
Keep
one storyboard, one request, the model’s own sound — no voice-over, no stock effects.
✓ Source experiment pick

One storyboard, one request

Thirteen 15-second episodes with Mara and her sidekick Pip, in thirteen styles — from claymation to ukiyo-e — each made from a single storyboard picture and a single request, with the voices and music the model made itself.

Input
one picture — a 3×3 storyboard, nine numbered frames, drawn by GPT Image 2.5 from your character sheet
Video
one MiniMax H3 Max Turbo request · 15 s · 480P · no end frame, no audio in
Sound
native — the model voices the lines, adds the foley and the music of the style
Time
15 s episode · 3–6 s of generation
See all 13 →
Reference guide

toon — a reference guide linking the steps

For your character, style and story, this external reference guide covers how to draw the nine-frame storyboard, write the cartoon prompt with the acting and the lines, generate the episode with its own voice and music, and trim it — with the exact settings and prompts from this page.

⬇Download reference guidetoonA reference guide connecting the creative steps — storyboard · one request · trim
  1. 1toon-storyboardNine frames of the story in one picture
  2. 2toon-turboOne H3 Max Turbo request with native sound

The short answer

A short cartoon does not need a pipeline. Draw the whole story as one storyboard picture — nine frames, your character, the style — and give it to MiniMax H3 Max Turbo with a detailed, emotional prompt: who does what in each shot, how they feel, what they say. The model animates it, voices both characters and scores it, in a few seconds in the source experiment.

  • Input✕ two keyframes for every shot✓ one 3×3 storyboard
  • Requests✕ one generation per shot, cut together✓ one generation per episode
  • Sound✕ text-to-speech and stock effects mixed in✓ the model’s native voices, foley and music
  • End frame✕ the character sheet as the last image✓ nothing — the board alone
  • Prompt✕ “the action moves from the first frame to the last”✓ nine shots with emotions, gags and lines

Draft or quality: Seedance 2.5

The external method on this page uses MiniMax H3 for drafts and many takes. Seedance 2.5 reference-to-video accepts the same pictures and sound as references (@Image1, @Audio1), up to 30 seconds a call, with its own synced sound at 480p, 720p or 1080p. The source preferred Seedance for quality. Side by side on our product commercial, its reported cuts were approximately one second early and half a second late; it drew finer detail and added its own sound effects in sync. This side-by-side result is from a product commercial, not a measurement on every format.

1 · The characters

GPT Image 2.5 Flare edit

Mara is our recurring character; for the cartoons she got a sidekick, Pip — a small red-orange fox-raccoon with a striped tail, invented in words. A sheet per style shows that one photo-real character sheet turns into any cartoon look; for the storyboards Mara’s own sheet is enough.

3D feature
3D feature
Anime
Anime
Claymation
Claymation

2 · The whole story in one picture

GPT Image 2.5 Flare edit · medium

The storyboard is the only input of the video: nine numbered 16:9 frames on black, three rows of three, one beat of the story each — setup, problem, gag, payoff. One GPT Image 2.5 request with Mara’s sheet as picture 1, the style in words and a sentence per frame.

Storyboard · Pip vs. the Vacuum · claymation · click for the prompt
Storyboard · Pip vs. the Vacuum · claymation · click for the prompt
Reference guides for this step · Markdown instructions for Claude Code or any agent

3 · One request with native sound

MiniMax H3 Max Turbo · image-to-video · 15 s · 480P

The board goes in as the first image — and nothing else: no last frame, no audio. The prompt does four things: says the board is not a shot and never appears on screen; describes the characters and their voices in words; walks through the nine shots with their seconds, the acting and the lines in quotes; and asks for the sound “generated with the picture”. Prompt enhance is off.

What came back across thirteen episodes: the board flashes for 1–7 frames at the very start (we trim it) and never returns; every line is spoken, roughly on time, in two distinct voices; the music fits the style — a 1930s jazz band, a chiptune, a koto, a synthwave track.

Prompt · Pip vs. the Vacuum
A 15-second stop-motion claymation comedy: plasticine and felt puppets with visible fingerprints, a handmade miniature living room, warm daylight, slightly stepped stop-motion movement, big rubbery squash-and-stretch gags and exaggerated faces. 16:9, nine snappy shots with hard cuts.

The given first image is NOT a shot: it is the storyboard for this film, nine numbered frames in three rows of three on black. It never appears on screen; the film starts directly inside frame 1, and every shot looks exactly like its storyboard frame enlarged to the full screen — the same characters, set, framing, colours and light. The screen is never split into a grid.

Characters: MARA, a young woman with black hair and an olive bomber jacket — a warm, low, playful voice. PIP, her tiny red-orange fox-raccoon sidekick with a striped bushy tail — a tiny, high, squeaky cartoon voice, fast and excitable.

Shot 1 (frame 1, 0–1.8 s): cleaning day — Mara in yellow rubber gloves stomps on the switch of a big old vacuum cleaner with a triumphant grin: "Cleaning day!" The vacuum roars to life.
Shot 2 (frame 2, 1.8–3.3 s): on the sofa, Pip's ears shoot straight up and every hair on his body stands on end, eyes bulging — a squeaky gasp.
Shot 3 (frame 3, 3.3–4.8 s): the roaring nozzle sweeps across the rug toward the sofa, Pip trembling in the background.
Shot 4 (frame 4, 4.8–6.3 s): Pip clings to the curtain with all four paws, screaming: "No no no no!"
Shot 5 (frame 5, 6.3–8 s): the nozzle catches his striped tail — SHLURP — and his whole body stretches long like rubber toward it, paws scrabbling.
Shot 6 (frame 6, 8–9.3 s): POP! He's gone. The vacuum chokes and its cloth dust bag balloons up huge and wobbles like a jelly.
Shot 7 (frame 7, 9.3–10.8 s): Mara, horrified, kills the switch and frantically unzips the bag: "Pip?!"
Shot 8 (frame 8, 10.8–12.8 s): Pip pops out, puffed up perfectly round and fluffy like a dandelion, dust drifting off him; he blinks once and says, deadpan: "...I'm clean."
Shot 9 (frame 9, 12.8–15 s): Mara collapses on the floor laughing helplessly while fluffy Pip sits with his arms crossed, utterly unamused, a single dust puff drifting off his head.

Sound: generated with the picture — the voices, the vacuum's roar and choke, the slurp and pop, Mara's laughter, light comedic foley for every move, a bouncy tuba-and-pizzicato comedy tune underneath; everything quieter than the voices. Each character speaks only their own lines, lips in sync. No subtitles, no text, no live action.
Reference guides for this step · Markdown instructions for Claude Code or any agent

4 · Thirteen styles, thirteen stories

each from one storyboard and one request
Ten of them at once, muted — each below with its own sound

1. The Last Donut 3D feature

Mara sneaks to the fridge at midnight for the last donut — and finds Pip on the top shelf, sugar on his whiskers.

Input
one storyboard picture
Model
MiniMax H3 Max Turbo · 15 s · 480P
Sound
native — voices, foley, music
Time
2.5 s
Storyboard and prompt
The only input · click for its prompt
storyboard
A 15-second animated comedy short in the style of a modern 3D animated feature film: soft subsurface skin, big expressive eyes, warm cinematic light, squash and stretch, snappy cartoon timing, broad heartfelt acting. 16:9, six quick shots with hard cuts.

The given first image is NOT a shot: it is the storyboard for this film, nine numbered frames in three rows of three on black. It never appears on screen; the film starts directly inside frame 1, and every shot looks exactly like its storyboard frames enlarged to the full screen — the same characters, kitchen, framing, colours and moonlit night light. The screen is never split into a grid.

Characters: MARA, a young woman with black hair and an olive bomber jacket — a warm, low, playful voice. PIP, her tiny red-orange fox-raccoon sidekick with a striped bushy tail — a tiny, high, squeaky cartoon voice, fast and nervous.

Shot 1 (frames 1–2, 0–3 s): a dark kitchen at midnight, moonlight through the window, a clock ticking. Mara tiptoes across the floor in exaggerated sneaky steps, knees high, shoulders hunched, flashlight beam bobbing ahead of her. She reaches the fridge, licks her lips and whispers with delicious anticipation: "Finally... the last donut."
Shot 2 (frames 3–4, 3–5 s): the fridge door swings open with a soft pop and a hum; golden fridge light floods her face — eyes huge and sparkling, a dreamy smile. Then a crash-zoom onto the plate: empty. Only crumbs and a dusting of powdered sugar. A sad little silence.
Shot 3 (frames 5–6, 5–8 s): on the top shelf, Pip freezes mid-chew, cheeks stuffed, powdered sugar all over his whiskers and nose, eyes wide like saucers. He gulps, then slaps on a huge fake innocent smile, paws hidden behind his back, tail swishing guiltily, and squeaks: "It wasn't me!"
Shot 4 (frame 7, 8–10 s): a close-up of Mara lit from below by the fridge. Her smile drops; one eyebrow rises slowly; her eyes narrow to razor-thin suspicious slits. A long beat. Flat and slow: "Pip."
Shot 5 (frame 8, 10–12 s): Pip's cheeks puff up, his eyes cross — HIC! A tiny squeaky hiccup, and a perfect little cloud of powdered sugar puffs out of his mouth and hangs in the air. He looks at it in horror, ears flattening.
Shot 6 (frame 9, 12–15 s): the wide kitchen in warm fridge light. Mara bursts out laughing, doubled over, holding her belly, wiping a tear: "Oh my God, Pip!" Pip giggles sheepishly, scratching the back of his head: "Hehe... sorry." A tiny playful music sting ends the film.

Sound: generated with the picture — the voices, the ticking clock, the fridge pop and hum, Pip's hiccup, soft cartoon foley for every move, all quieter than the voices; no music until the final sting. Each character speaks only their own lines, lips in sync. No subtitles, no text, no live action.

2. First Snow anime

Pip sees snow for the first time, chases the flakes, sneezes himself into a snowdrift and ends up in Mara’s scarf.

Input
one storyboard picture
Model
MiniMax H3 Max Turbo · 15 s · 480P
Sound
native — voices, foley, music
Time
3.7 s
Storyboard and prompt
The only input · click for its prompt
storyboard
A 15-second animated short in the style of a 2D Japanese anime feature film: clean line art, cel shading, hand-painted backgrounds, soft snowy night light, tender heartfelt acting with big anime emotions — sparkling eyes, blushing cheeks, sweat-drop and sparkle effects drawn in the anime way. 16:9, nine shots with gentle cuts.

The given first image is NOT a shot: it is the storyboard for this film, nine numbered frames in three rows of three on black. It never appears on screen; the film starts directly inside frame 1, and every shot looks exactly like its storyboard frame enlarged to the full screen — the same characters, set, framing, colours and light. The screen is never split into a grid.

Characters: MARA, a young woman with black hair and an olive bomber jacket — a warm, low, playful voice. PIP, her tiny red-orange fox-raccoon sidekick with a striped bushy tail — a tiny, high, squeaky cartoon voice, fast and excitable.

Shot 1 (frame 1, 0–2 s): night; the first snowflakes drift past the window. Pip squashes his nose and paws against the glass, eyes shimmering, tail wagging wildly: "Mara! The sky is falling!"
Shot 2 (frame 2, 2–3.5 s): Mara on the sofa with a warm mug, her face softening into a tender smile: "It's snow, silly."
Shot 3 (frame 3, 3.5–5 s): the snowy balcony under a glowing streetlamp — Pip bursts out and leaps high into the air, paws stretched toward the flakes, Mara giggling in the doorway.
Shot 4 (frame 4, 5–6.5 s): Pip spins in happy circles in the fresh snow, tongue out, chasing flakes, squeaking with joy: "Wheee!"
Shot 5 (frame 5, 6.5–8 s): extreme close-up — one big perfect snowflake lands on the tip of Pip's nose; his eyes slowly cross to look at it, a tiny gasp.
Shot 6 (frame 6, 8–9 s): his nose twitches, his face scrunches up — "Ah... ah..."
Shot 7 (frame 7, 9–10.5 s): "ACHOO!" — a huge cartoon sneeze blasts him backwards into a soft snowdrift in an explosion of powder.
Shot 8 (frame 8, 10.5–12.5 s): Mara kneels and scoops the snow-covered Pip into her hands; he shakes himself off, both laughing.
Shot 9 (frame 9, 12.5–15 s): wrapped together in her long orange scarf, they watch the snow fall over the city lights; Mara whispers, warm and soft: "Happy first snow, Pip." Pip snuggles closer with a tiny happy sigh.

Sound: generated with the picture — the voices, the hush of falling snow, soft crunching steps, the sneeze, a quiet gentle piano and music-box melody that swells at the end; everything quieter than the voices. Each character speaks only their own lines, lips in sync. No subtitles, no text, no live action.

3. Pip vs. the Vacuum claymation

Cleaning day: the vacuum swallows Pip by the tail, the dust bag balloons, and he comes out round and fluffy as a dandelion.

Input
one storyboard picture
Model
MiniMax H3 Max Turbo · 15 s · 480P
Sound
native — voices, foley, music
Time
2.5 s
Storyboard and prompt
The only input · click for its prompt
storyboard
A 15-second stop-motion claymation comedy: plasticine and felt puppets with visible fingerprints, a handmade miniature living room, warm daylight, slightly stepped stop-motion movement, big rubbery squash-and-stretch gags and exaggerated faces. 16:9, nine snappy shots with hard cuts.

The given first image is NOT a shot: it is the storyboard for this film, nine numbered frames in three rows of three on black. It never appears on screen; the film starts directly inside frame 1, and every shot looks exactly like its storyboard frame enlarged to the full screen — the same characters, set, framing, colours and light. The screen is never split into a grid.

Characters: MARA, a young woman with black hair and an olive bomber jacket — a warm, low, playful voice. PIP, her tiny red-orange fox-raccoon sidekick with a striped bushy tail — a tiny, high, squeaky cartoon voice, fast and excitable.

Shot 1 (frame 1, 0–1.8 s): cleaning day — Mara in yellow rubber gloves stomps on the switch of a big old vacuum cleaner with a triumphant grin: "Cleaning day!" The vacuum roars to life.
Shot 2 (frame 2, 1.8–3.3 s): on the sofa, Pip's ears shoot straight up and every hair on his body stands on end, eyes bulging — a squeaky gasp.
Shot 3 (frame 3, 3.3–4.8 s): the roaring nozzle sweeps across the rug toward the sofa, Pip trembling in the background.
Shot 4 (frame 4, 4.8–6.3 s): Pip clings to the curtain with all four paws, screaming: "No no no no!"
Shot 5 (frame 5, 6.3–8 s): the nozzle catches his striped tail — SHLURP — and his whole body stretches long like rubber toward it, paws scrabbling.
Shot 6 (frame 6, 8–9.3 s): POP! He's gone. The vacuum chokes and its cloth dust bag balloons up huge and wobbles like a jelly.
Shot 7 (frame 7, 9.3–10.8 s): Mara, horrified, kills the switch and frantically unzips the bag: "Pip?!"
Shot 8 (frame 8, 10.8–12.8 s): Pip pops out, puffed up perfectly round and fluffy like a dandelion, dust drifting off him; he blinks once and says, deadpan: "...I'm clean."
Shot 9 (frame 9, 12.8–15 s): Mara collapses on the floor laughing helplessly while fluffy Pip sits with his arms crossed, utterly unamused, a single dust puff drifting off his head.

Sound: generated with the picture — the voices, the vacuum's roar and choke, the slurp and pop, Mara's laughter, light comedic foley for every move, a bouncy tuba-and-pizzicato comedy tune underneath; everything quieter than the voices. Each character speaks only their own lines, lips in sync. No subtitles, no text, no live action.

4. The Lost Balloon watercolour picture book

Pip lets go of his red balloon — and watches it float to a little girl on a balcony.

Input
one storyboard picture
Model
MiniMax H3 Max Turbo · 15 s · 480P
Sound
native — voices, foley, music
Time
6.2 s
Storyboard and prompt
The only input · click for its prompt
storyboard
A 15-second animated short, "The Lost Balloon", in the style of a children's picture-book watercolour animation: soft wet-on-wet washes, pencil outlines, paper texture, gentle pastel light. Expressive, funny, heartfelt cartoon acting: big emotions on every face, squash and stretch, snappy comic timing. 16:9, nine shots with cuts at the given times.

The given first image is NOT a shot: it is the storyboard for this film, nine numbered frames in three rows of three on black. It never appears on screen; the film starts directly inside frame 1, and every shot looks exactly like its storyboard frame enlarged to the full screen — the same characters, set, framing, colours and light. The screen is never split into a grid.

Characters: MARA, a young woman with black hair and an olive bomber jacket — a warm, low, playful voice. PIP — her tiny red-orange fox-raccoon sidekick with a black raccoon mask around big eyes and a striped bushy tail — a tiny, high, squeaky cartoon voice, fast and excitable.

Shot 1 (frame 1, 0.0–1.7 s): a sunny park; Pip holds a big red balloon on a string, beaming. Pip: "Look, Mara! It's mine!"
Shot 2 (frame 2, 1.7–3.3 s): a gust of wind tugs the balloon; Pip's feet lift off the grass. Pip: "Whoa-whoa-whoa!"
Shot 3 (frame 3, 3.3–5.0 s): Pip lets go; the balloon floats up into the sky.
Shot 4 (frame 4, 5.0–6.6 s): Pip's lip trembles, big tears welling in his eyes. Pip, sniffling: "My balloon..."
Shot 5 (frame 5, 6.6–8.3 s): Mara kneels beside him and points up with a soft smile. Mara: "Look where it's going."
Shot 6 (frame 6, 8.3–9.9 s): the balloon drifts over rooftops toward a little girl on a balcony.
Shot 7 (frame 7, 9.9–11.6 s): the girl catches the string and gasps with delight.
Shot 8 (frame 8, 11.6–13.2 s): the girl waves down at them; Pip slowly smiles. Pip: "...She looks happy."
Shot 9 (frame 9, 13.2–15.0 s): Mara and Pip walk home hand in paw at sunset. Mara: "So are you."

Sound: generated with the picture — the voices, birdsong, a soft breeze, a gentle acoustic guitar and glockenspiel lullaby, and cartoon foley for every action, all quieter than the voices. Each character speaks only their own lines, lips in sync. No subtitles, no text, no live action.

5. Pip's Piano Recital 1930s rubber hose

Pip’s piano recital goes wrong when the piano starts biting and chases him round the stage.

Input
one storyboard picture
Model
MiniMax H3 Max Turbo · 15 s · 480P
Sound
native — voices, foley, music
Time
3.7 s
Storyboard and prompt
The only input · click for its prompt
storyboard
A 15-second animated short, "Pip's Piano Recital", in the style of a 1930s black-and-white rubber-hose cartoon: bouncy noodle limbs, pie-cut eyes, white gloves, film grain, flicker and dust, everything bobbing to the beat. Expressive, funny, heartfelt cartoon acting: big emotions on every face, squash and stretch, snappy comic timing. 16:9, nine shots with cuts at the given times.

The given first image is NOT a shot: it is the storyboard for this film, nine numbered frames in three rows of three on black. It never appears on screen; the film starts directly inside frame 1, and every shot looks exactly like its storyboard frame enlarged to the full screen — the same characters, set, framing, colours and light. The screen is never split into a grid.

Characters: MARA, a young woman with black hair and an olive bomber jacket — a warm, low, playful voice. PIP — her tiny red-orange fox-raccoon sidekick with a black raccoon mask around big eyes and a striped bushy tail — a tiny, high, squeaky cartoon voice, fast and excitable.

Shot 1 (frame 1, 0.0–1.7 s): a little theatre stage; Mara in the front row claps. Mara: "Go, Pip!"
Shot 2 (frame 2, 1.7–3.3 s): Pip bows with a flourish and sits at a tiny grand piano.
Shot 3 (frame 3, 3.3–5.0 s): he cracks his knuckles, fingers stretching like noodles.
Shot 4 (frame 4, 5.0–6.6 s): he plays a jaunty tune, the piano bouncing to the beat.
Shot 5 (frame 5, 6.6–8.3 s): the piano keys start biting at his fingers like teeth. Pip: "Hey!"
Shot 6 (frame 6, 8.3–9.9 s): the piano chases Pip around the stage on its little legs. Pip: "Help!"
Shot 7 (frame 7, 9.9–11.6 s): Pip leaps into Mara's arms.
Shot 8 (frame 8, 11.6–13.2 s): the piano sits down, panting like a tired dog.
Shot 9 (frame 9, 13.2–15.0 s): the audience roars; Mara and Pip bow together. Mara: "Encore!"

Sound: generated with the picture — the voices, a scratchy 1930s jazz band, piano, slide whistles, honk horns and cartoon boings, and cartoon foley for every action, all quieter than the voices. Each character speaks only their own lines, lips in sync. No subtitles, no text, no live action.

6. Level Up 16-bit pixel art

A side-scrolling level: springs, coins, a slime boss and one tiny donut in the chest.

Input
one storyboard picture
Model
MiniMax H3 Max Turbo · 15 s · 480P
Sound
native — voices, foley, music
Time
3.8 s
Storyboard and prompt
The only input · click for its prompt
storyboard
A 15-second animated short, "Level Up", in the style of a 16-bit pixel-art video game: chunky pixels, a limited bright palette, side-scrolling parallax backgrounds, sprite animation. Expressive, funny, heartfelt cartoon acting: big emotions on every face, squash and stretch, snappy comic timing. 16:9, nine shots with cuts at the given times.

The given first image is NOT a shot: it is the storyboard for this film, nine numbered frames in three rows of three on black. It never appears on screen; the film starts directly inside frame 1, and every shot looks exactly like its storyboard frame enlarged to the full screen — the same characters, set, framing, colours and light. The screen is never split into a grid.

Characters: MARA, a young woman with black hair and an olive bomber jacket — a warm, low, playful voice. PIP — her tiny red-orange fox-raccoon sidekick with a black raccoon mask around big eyes and a striped bushy tail — a tiny, high, squeaky cartoon voice, fast and excitable.

Shot 1 (frame 1, 0.0–1.7 s): a title-screen sky; Mara and Pip stand at the start of a green grassy level. Mara: "Ready, Pip?"
Shot 2 (frame 2, 1.7–3.3 s): they run right; Pip bounces on a mushroom springboard. Pip: "Boing!"
Shot 3 (frame 3, 3.3–5.0 s): Mara jumps over a pit, grabbing a spinning gold coin.
Shot 4 (frame 4, 5.0–6.6 s): Pip collects a row of coins in a rapid shower of sparkles.
Shot 5 (frame 5, 6.6–8.3 s): a giant pixel slime boss drops down, wobbling. Pip: "Uh oh."
Shot 6 (frame 6, 8.3–9.9 s): Mara throws Pip like a ball; he bounces off the boss's head.
Shot 7 (frame 7, 9.9–11.6 s): the boss bursts into pixel stars.
Shot 8 (frame 8, 11.6–13.2 s): a big treasure chest opens: one tiny donut inside. Pip: "A donut!"
Shot 9 (frame 9, 13.2–15.0 s): they cheer, frozen in a victory pose under falling confetti. Mara: "Level complete!"

Sound: generated with the picture — the voices, a catchy 16-bit chiptune, jump blips, coin dings and an 8-bit victory fanfare, and cartoon foley for every action, all quieter than the voices. Each character speaks only their own lines, lips in sync. No subtitles, no text, no live action.

7. Rooftop Chase comic book

Pip steals Mara’s keys and leads her on a chase across the rooftops at sunset.

Input
one storyboard picture
Model
MiniMax H3 Max Turbo · 15 s · 480P
Sound
native — voices, foley, music
Time
5.0 s
Storyboard and prompt
The only input · click for its prompt
storyboard
A 15-second animated short, "Rooftop Chase", in the style of a comic-book animation: bold ink lines, Ben-Day halftone dots, vivid flat colours, off-register printing, dynamic comic panel energy and speed lines. Expressive, funny, heartfelt cartoon acting: big emotions on every face, squash and stretch, snappy comic timing. 16:9, nine shots with cuts at the given times.

The given first image is NOT a shot: it is the storyboard for this film, nine numbered frames in three rows of three on black. It never appears on screen; the film starts directly inside frame 1, and every shot looks exactly like its storyboard frame enlarged to the full screen — the same characters, set, framing, colours and light. The screen is never split into a grid.

Characters: MARA, a young woman with black hair and an olive bomber jacket — a warm, low, playful voice. PIP — her tiny red-orange fox-raccoon sidekick with a black raccoon mask around big eyes and a striped bushy tail — a tiny, high, squeaky cartoon voice, fast and excitable.

Shot 1 (frame 1, 0.0–1.7 s): a city rooftop at dusk; Pip snatches Mara's keys and grins. Pip: "Catch me if you can!"
Shot 2 (frame 2, 1.7–3.3 s): Pip leaps across a gap between buildings, speed lines behind him.
Shot 3 (frame 3, 3.3–5.0 s): Mara sprints after him, jacket flapping. Mara: "Pip! Those are my keys!"
Shot 4 (frame 4, 5.0–6.6 s): Pip slides down a fire escape rail.
Shot 5 (frame 5, 6.6–8.3 s): Mara vaults over a chimney in slow heroic action.
Shot 6 (frame 6, 8.3–9.9 s): Pip stops at the roof edge, wobbling, arms windmilling. Pip: "Whoa!"
Shot 7 (frame 7, 9.9–11.6 s): Mara grabs him by the tail just in time.
Shot 8 (frame 8, 11.6–13.2 s): she holds him upside down; the keys jingle out of his paws. Mara: "Gotcha."
Shot 9 (frame 9, 13.2–15.0 s): they sit on the edge watching the sunset, Pip sheepish. Pip: "Rematch tomorrow?"

Sound: generated with the picture — the voices, a punchy action-cartoon orchestra, whooshes, footsteps and city wind, and cartoon foley for every action, all quieter than the voices. Each character speaks only their own lines, lips in sync. No subtitles, no text, no live action.

8. The Magic Seed paper cut-out

A single seed grows into a sunflower that sneezes its seeds all over Pip.

Input
one storyboard picture
Model
MiniMax H3 Max Turbo · 15 s · 480P
Sound
native — voices, foley, music
Time
3.8 s
Storyboard and prompt
The only input · click for its prompt
storyboard
A 15-second animated short, "The Magic Seed", in the style of a paper cut-out stop-motion animation: layered coloured paper and card, visible paper edges and shadows between layers, slightly jerky movement. Expressive, funny, heartfelt cartoon acting: big emotions on every face, squash and stretch, snappy comic timing. 16:9, nine shots with cuts at the given times.

The given first image is NOT a shot: it is the storyboard for this film, nine numbered frames in three rows of three on black. It never appears on screen; the film starts directly inside frame 1, and every shot looks exactly like its storyboard frame enlarged to the full screen — the same characters, set, framing, colours and light. The screen is never split into a grid.

Characters: MARA, a young woman with black hair and an olive bomber jacket — a warm, low, playful voice. PIP — her tiny red-orange fox-raccoon sidekick with a black raccoon mask around big eyes and a striped bushy tail — a tiny, high, squeaky cartoon voice, fast and excitable.

Shot 1 (frame 1, 0.0–1.7 s): a spring garden; Mara hands Pip a single tiny seed. Mara: "Plant it, and wait."
Shot 2 (frame 2, 1.7–3.3 s): Pip digs a hole with his paws and drops the seed in.
Shot 3 (frame 3, 3.3–5.0 s): he waters it with a big watering can, splashing himself.
Shot 4 (frame 4, 5.0–6.6 s): he sits and stares at the dirt, tapping his foot. Pip: "Is it done yet?"
Shot 5 (frame 5, 6.6–8.3 s): a tiny green sprout pops up. Pip: "Ooh!"
Shot 6 (frame 6, 8.3–9.9 s): the sprout grows and grows, shooting upward.
Shot 7 (frame 7, 9.9–11.6 s): a giant sunflower towers over them, blocking the sun.
Shot 8 (frame 8, 11.6–13.2 s): the sunflower bends down and sneezes seeds all over Pip.
Shot 9 (frame 9, 13.2–15.0 s): Pip, covered in seeds, grins at Mara. Pip: "I'm gonna need more holes."

Sound: generated with the picture — the voices, paper rustles, birdsong, a cheerful ukulele and xylophone tune, and cartoon foley for every action, all quieter than the voices. Each character speaks only their own lines, lips in sync. No subtitles, no text, no live action.

9. Rocket to the Moon flat vector

A cardboard rocket on soda fizz takes Pip to the moon — and the moon blows him home.

Input
one storyboard picture
Model
MiniMax H3 Max Turbo · 15 s · 480P
Sound
native — voices, foley, music
Time
3.7 s
Storyboard and prompt
The only input · click for its prompt
storyboard
A 15-second animated short, "Rocket to the Moon", in the style of a flat vector motion-graphics animation: clean geometric shapes, bold flat colours, soft gradients, smooth easing, an explainer-video look. Expressive, funny, heartfelt cartoon acting: big emotions on every face, squash and stretch, snappy comic timing. 16:9, nine shots with cuts at the given times.

The given first image is NOT a shot: it is the storyboard for this film, nine numbered frames in three rows of three on black. It never appears on screen; the film starts directly inside frame 1, and every shot looks exactly like its storyboard frame enlarged to the full screen — the same characters, set, framing, colours and light. The screen is never split into a grid.

Characters: MARA, a young woman with black hair and an olive bomber jacket — a warm, low, playful voice. PIP — her tiny red-orange fox-raccoon sidekick with a black raccoon mask around big eyes and a striped bushy tail — a tiny, high, squeaky cartoon voice, fast and excitable.

Shot 1 (frame 1, 0.0–1.7 s): a starry backyard; Mara and Pip finish a cardboard rocket. Pip: "Is it ready?"
Shot 2 (frame 2, 1.7–3.3 s): Pip climbs in wearing a colander helmet.
Shot 3 (frame 3, 3.3–5.0 s): countdown: Mara holds up three fingers. Mara: "Three, two, one..."
Shot 4 (frame 4, 5.0–6.6 s): the rocket blasts off on a fountain of fizzing soda.
Shot 5 (frame 5, 6.6–8.3 s): the rocket arcs past clouds and satellites. Pip: "Wheeee!"
Shot 6 (frame 6, 8.3–9.9 s): Pip plants a little flag on the moon.
Shot 7 (frame 7, 9.9–11.6 s): the moon has a face; it yawns.
Shot 8 (frame 8, 11.6–13.2 s): the moon gently blows Pip back home on a breath.
Shot 9 (frame 9, 13.2–15.0 s): Pip lands in Mara's arms in the backyard. Pip: "Again!"

Sound: generated with the picture — the voices, an upbeat synth explainer tune, whooshes, soft pops and twinkles, and cartoon foley for every action, all quieter than the voices. Each character speaks only their own lines, lips in sync. No subtitles, no text, no live action.

10. Winter Fishing 1970s Soviet hand-drawn

Ice fishing at dawn: the catch is an old boot, with a winking fish inside.

Input
one storyboard picture
Model
MiniMax H3 Max Turbo · 15 s · 480P
Sound
native — voices, foley, music
Time
3.7 s
Storyboard and prompt
The only input · click for its prompt
storyboard
A 15-second animated short, "Winter Fishing", in the style of a 1970s Soviet hand-drawn animation: soft painted backgrounds, gentle line, muted warm colours, cosy winter fairy-tale charm. Expressive, funny, heartfelt cartoon acting: big emotions on every face, squash and stretch, snappy comic timing. 16:9, nine shots with cuts at the given times.

The given first image is NOT a shot: it is the storyboard for this film, nine numbered frames in three rows of three on black. It never appears on screen; the film starts directly inside frame 1, and every shot looks exactly like its storyboard frame enlarged to the full screen — the same characters, set, framing, colours and light. The screen is never split into a grid.

Characters: MARA, a young woman with black hair and an olive bomber jacket — a warm, low, playful voice. PIP — her tiny red-orange fox-raccoon sidekick with a black raccoon mask around big eyes and a striped bushy tail — a tiny, high, squeaky cartoon voice, fast and excitable.

Shot 1 (frame 1, 0.0–1.7 s): a frozen lake at dawn; Mara and Pip sit by a hole in the ice with tiny fishing rods. Mara: "Patience, Pip."
Shot 2 (frame 2, 1.7–3.3 s): Pip stares at the hole, fidgeting.
Shot 3 (frame 3, 3.3–5.0 s): the float bobs; Pip's eyes go wide. Pip: "A bite!"
Shot 4 (frame 4, 5.0–6.6 s): Pip pulls with all his might, sliding on the ice.
Shot 5 (frame 5, 6.6–8.3 s): Mara grabs his tail and pulls too.
Shot 6 (frame 6, 8.3–9.9 s): out flies... an old boot.
Shot 7 (frame 7, 9.9–11.6 s): Pip stares at the boot, deflated. Pip: "A boot."
Shot 8 (frame 8, 11.6–13.2 s): a tiny fish peeks out of the boot and winks.
Shot 9 (frame 9, 13.2–15.0 s): they laugh; the fish dives back into the hole. Mara: "Let it go."

Sound: generated with the picture — the voices, a quiet frosty morning, crunching ice, a gentle balalaika and flute melody, and cartoon foley for every action, all quieter than the voices. Each character speaks only their own lines, lips in sync. No subtitles, no text, no live action.

11. Picnic and the Wind ukiyo-e woodblock

A great gust steals the picnic; Pip catches his rice ball over a wave and rides a seagull back.

Input
one storyboard picture
Model
MiniMax H3 Max Turbo · 15 s · 480P
Sound
native — voices, foley, music
Time
3.7 s
Storyboard and prompt
The only input · click for its prompt
storyboard
A 15-second animated short, "Picnic and the Wind", in the style of an animated Japanese ukiyo-e woodblock print: flat colour blocks, fine carved outlines, washi paper texture, stylised waves and clouds. Expressive, funny, heartfelt cartoon acting: big emotions on every face, squash and stretch, snappy comic timing. 16:9, nine shots with cuts at the given times.

The given first image is NOT a shot: it is the storyboard for this film, nine numbered frames in three rows of three on black. It never appears on screen; the film starts directly inside frame 1, and every shot looks exactly like its storyboard frame enlarged to the full screen — the same characters, set, framing, colours and light. The screen is never split into a grid.

Characters: MARA, a young woman with black hair and an olive bomber jacket — a warm, low, playful voice. PIP — her tiny red-orange fox-raccoon sidekick with a black raccoon mask around big eyes and a striped bushy tail — a tiny, high, squeaky cartoon voice, fast and excitable.

Shot 1 (frame 1, 0.0–1.7 s): a hillside by the sea; Mara and Pip spread a picnic under a pine tree. Mara: "Perfect day."
Shot 2 (frame 2, 1.7–3.3 s): Pip unwraps a rice ball, delighted.
Shot 3 (frame 3, 3.3–5.0 s): a great curling gust of wind sweeps in, stylised swirls.
Shot 4 (frame 4, 5.0–6.6 s): the picnic cloth flies up, plates spinning in the air. Pip: "My lunch!"
Shot 5 (frame 5, 6.6–8.3 s): Pip leaps after the rice ball through the air.
Shot 6 (frame 6, 8.3–9.9 s): he catches it mid-air above a stylised wave.
Shot 7 (frame 7, 9.9–11.6 s): he lands on a passing seagull's back. Pip: "Whoa!"
Shot 8 (frame 8, 11.6–13.2 s): the gull circles and drops him back on the blanket.
Shot 9 (frame 9, 13.2–15.0 s): Pip bites the rice ball triumphantly; Mara laughs. Pip: "Worth it."

Sound: generated with the picture — the voices, wind chimes, waves, a koto and shakuhachi melody, and cartoon foley for every action, all quieter than the voices. Each character speaks only their own lines, lips in sync. No subtitles, no text, no live action.

12. The Paper Plane chalkboard

A paper plane drawn in chalk flies Pip around the Eiffel Tower, through rain and a rainbow.

Input
one storyboard picture
Model
MiniMax H3 Max Turbo · 15 s · 480P
Sound
native — voices, foley, music
Time
2.5 s
Storyboard and prompt
The only input · click for its prompt
storyboard
A 15-second animated short, "The Paper Plane", in the style of a chalkboard animation: white and coloured chalk lines drawn on a dark green blackboard, smudges and dust, lines that draw themselves. Expressive, funny, heartfelt cartoon acting: big emotions on every face, squash and stretch, snappy comic timing. 16:9, nine shots with cuts at the given times.

The given first image is NOT a shot: it is the storyboard for this film, nine numbered frames in three rows of three on black. It never appears on screen; the film starts directly inside frame 1, and every shot looks exactly like its storyboard frame enlarged to the full screen — the same characters, set, framing, colours and light. The screen is never split into a grid.

Characters: MARA, a young woman with black hair and an olive bomber jacket — a warm, low, playful voice. PIP — her tiny red-orange fox-raccoon sidekick with a black raccoon mask around big eyes and a striped bushy tail — a tiny, high, squeaky cartoon voice, fast and excitable.

Shot 1 (frame 1, 0.0–1.7 s): a chalk classroom; Mara folds a paper plane and hands it to Pip. Mara: "Make a wish."
Shot 2 (frame 2, 1.7–3.3 s): Pip throws it out of the window. Pip: "Fly!"
Shot 3 (frame 3, 3.3–5.0 s): the plane swoops over chalk rooftops.
Shot 4 (frame 4, 5.0–6.6 s): Pip leaps and rides on top of the plane.
Shot 5 (frame 5, 6.6–8.3 s): they loop around a chalk Eiffel tower. Pip: "Woo-hoo!"
Shot 6 (frame 6, 8.3–9.9 s): they fly through a chalk rainstorm.
Shot 7 (frame 7, 9.9–11.6 s): a chalk sun draws itself and the rain clears into a rainbow.
Shot 8 (frame 8, 11.6–13.2 s): the plane glides back through the classroom window.
Shot 9 (frame 9, 13.2–15.0 s): Pip lands on Mara's desk, chalk dust everywhere. Pip: "Where to tomorrow?"

Sound: generated with the picture — the voices, chalk scratching, wind, a playful whistled tune and piano, and cartoon foley for every action, all quieter than the voices. Each character speaks only their own lines, lips in sync. No subtitles, no text, no live action.

13. Arcade Night neon synthwave

A dance-off on an arcade machine: PERFECT, a breakdance spin and neon fireworks.

Input
one storyboard picture
Model
MiniMax H3 Max Turbo · 15 s · 480P
Sound
native — voices, foley, music
Time
2.5 s
Storyboard and prompt
The only input · click for its prompt
storyboard
A 15-second animated short, "Arcade Night", in the style of an 80s neon synthwave animation: glowing neon outlines, magenta and cyan, a grid horizon, VHS scanlines and chromatic glow. Expressive, funny, heartfelt cartoon acting: big emotions on every face, squash and stretch, snappy comic timing. 16:9, nine shots with cuts at the given times.

The given first image is NOT a shot: it is the storyboard for this film, nine numbered frames in three rows of three on black. It never appears on screen; the film starts directly inside frame 1, and every shot looks exactly like its storyboard frame enlarged to the full screen — the same characters, set, framing, colours and light. The screen is never split into a grid.

Characters: MARA, a young woman with black hair and an olive bomber jacket — a warm, low, playful voice. PIP — her tiny red-orange fox-raccoon sidekick with a black raccoon mask around big eyes and a striped bushy tail — a tiny, high, squeaky cartoon voice, fast and excitable.

Shot 1 (frame 1, 0.0–1.7 s): a neon arcade at night; Mara and Pip step onto a dance machine. Mara: "Dance-off!"
Shot 2 (frame 2, 1.7–3.3 s): glowing arrows fly up the screen.
Shot 3 (frame 3, 3.3–5.0 s): Mara stomps the arrows perfectly, hair flying.
Shot 4 (frame 4, 5.0–6.6 s): Pip copies her, his tiny paws a blur. Pip: "Too fast!"
Shot 5 (frame 5, 6.6–8.3 s): Pip spins on his tail like a breakdancer.
Shot 6 (frame 6, 8.3–9.9 s): the screen flashes PERFECT in neon.
Shot 7 (frame 7, 9.9–11.6 s): Pip slips and slides across the floor. Pip: "Whoa!"
Shot 8 (frame 8, 11.6–13.2 s): he lands in a heroic pose in a neon spotlight.
Shot 9 (frame 9, 13.2–15.0 s): they high-five; neon fireworks burst behind them. Pip: "New high score!"

Sound: generated with the picture — the voices, a driving retro synthwave track, arcade bleeps and crowd cheers, and cartoon foley for every action, all quieter than the voices. Each character speaks only their own lines, lips in sync. No subtitles, no text, no live action.

How we got here

the same episode three ways
✕How not toa generation per shot, ElevenLabs sound
  • Two keyframes for each of six shots, 18 Turbo requests in three styles, the voices and effects from ElevenLabs.
  • The effects came out louder than the voices and odd; the acting was stiffer. Dropped — no ElevenLabs for cartoons.
✓How toone storyboard, one request
  • The same story drawn as one 3×3 board, one Turbo request, the sound made by the model.
  • Livelier faces and both voices.

In between we tried a 12-frame board as the first image and the character sheet as the last: the sheet flashed at the end, and in one style a whole shot played over it. Without a last image the problem is gone.

Settings and measurements

fal
StepToolTime
Character sheet in a style (optional)GPT Image 2.5 Flare edit, low, 2048 × 115227 s
Storyboard, nine framesGPT Image 2.5 Flare edit, medium, 1920 × 1088~30 s
Episode, 15 s, with its soundMiniMax H3 Max Turbo image-to-video, 480P2.5–6 s
Thirteen episodes13 boards + 13 requests

What works best

our conclusions
✓ Source experiment pick

One storyboard, one request

Thirteen 15-second episodes with Mara and her sidekick Pip, in thirteen styles — from claymation to ukiyo-e — each made from a single storyboard picture and a single request, with the voices and music the model made itself.

Input
one picture — a 3×3 storyboard, nine numbered frames, drawn by GPT Image 2.5 from your character sheet
Video
one MiniMax H3 Max Turbo request · 15 s · 480P · no end frame, no audio in
Sound
native — the model voices the lines, adds the foley and the music of the style
Time
15 s episode · 3–6 s of generation
See all 13 →
  • Draw the story, then animate it once. One storyboard picture carries the characters, the style and every shot.
  • Write the acting, not the motion. Emotions, gags and timing per shot make the faces come alive.
  • Describe the voices in words and put every line in quotes with who says it — the model casts and voices them.
  • Let the model make the sound. Its voices, foley and music fit the picture better than anything mixed in afterwards.
  • No last image — and trim the one-to-seven-frame flash of the board at the start.
  • Any style holds — from claymation and anime to ukiyo-e and pixel art — when the board is drawn in it.
FlashBoardsMake your character’s cartoon

The storyboard and every episode on one board.

Open FlashBoards

FAQ

5 questions
How do I make an AI cartoon with my own character?

Draw a 3×3 storyboard of the episode in one GPT Image 2.5 request with your character sheet as the reference and the style in words, then give that one picture to MiniMax H3 Max Turbo image-to-video with a detailed prompt — nine shots with their seconds, the acting and the lines. 15 seconds come back with voices and music in a few seconds.

Can MiniMax H3 generate voices and music for a cartoon?

Yes. Without an audio input H3 Max Turbo generates the sound with the picture: it voiced both characters from a description of their voices, spoke every quoted line roughly on time, added foley and music that fits the style.

Which cartoon styles work?

All thirteen we tried held from a storyboard drawn in them: 3D feature animation, anime, claymation, watercolour picture book, 1930s rubber hose, 16-bit pixel art, comic book, paper cut-out, flat vector, 1970s Soviet hand-drawn, ukiyo-e, chalkboard and neon synthwave.

Why not generate the cartoon shot by shot?

We did: two keyframes per shot, one request each, with separate voice-over and sound effects. The acting was stiffer and the mixed-in sound came out odd. One storyboard and one request gave a livelier episode.

What inputs does a 15-second AI cartoon need?

One storyboard picture and one H3 Max Turbo request at 480P, with the sound generated natively.

3

Examples use external tools; available FlashBoards workflows may differ. Downloads are reference guides, not installed automation.