Online AI video creation tool

Pika H3 Generator

Create AI video from a text prompt or image with a browser-based MiniMax H3 workflow.

Definition

Pika H3 Generator is an online AI video creation tool that allows users to generate videos from text prompts and images using MiniMax H3 technology.

Create AI Video

Try MiniMax H3

Choose a workflow, add your prompt or image, review the settings, and generate your video.

Primary reference image0/1
Secondary reference image (optional)0/1
Reference images
0/9 total

First 5 images are included; each additional image uses 3 credits.

Reference videos
0/3

MP4 / MOV / MKV, up to 3 files, 50MB each

Reference audio
0/3

MP3 / WAV / AAC / FLAC, up to 3 files, 15MB each

0/2500
768pRecommended

Recommended for testing prompts and first drafts.

2K output is available on the 2K tier only.

5s
Est. cost: ~30 creditsBalance: — credits

Checking your account…

OutputVideo
History
MiniMax H3 AI video: Cyber-Grunge Fashion FilmMiniMax H3 Reference to Video sample

Pick a result, make it yours

Create a Video Like This

Choose an example and your prompt, model, and settings load instantly.

Vintage Binocular Brand Film AI video previewProduct Ads

Vintage Binocular Brand Film

Product Ads
Model
MiniMax H3
Duration
5 seconds
Resolution
768p
Aspect ratio
16:9

“Use Images 1–4 as sequential keyframes, seen through a vintage binocular viewfinder searching for the MINIMAX installation. Open out of focus with subtle handheld shake, then push in quickly and rack focus onto Image 1. Between keyframes, use fast binocular-scan transitions with whip movement, motion blur, optical smearing, and brief exposure flicker. Cut at peak blur, then settle and snap back into focus. Keep the twin circular lens mask absolutely fixed throughout: identical position, scale, feathered black vignette, and edge softness, with no warping or drift. Only the image inside the mask may move. In Image 2, let the fabric move gently in the wind while the MINIMAX lettering follows the folds and remains legible. In Image 3, the subject should feel like a stylish passerby caught by chance, walking, turning, and swinging their arms naturally. In Image 4, the subject adjusts their glasses or lifts their chin slightly with a cool, effortless fashion-campaign attitude. Red typography should resolve with the focus: begin slightly blurred and at low opacity, then fade into clarity over 0.3–0.5 seconds. A subtle vertical slide or slight tracking expansion is allowed. Fade it out before the next transition or let motion blur carry it away. No spins, bounces, or large fly-ins/outs. Visual language: a voyeuristic, Wes Anderson-inspired 35 mm film look with fine grain, soft highlight halation, restrained color, and red typographic accents. Minimal, premium, lightly playful. Do not add people, vehicles, buildings, or logos. Preserve the core composition and the MINIMAX installation exactly. PRODUCTION CONTROL MODE: Reference to Video. Assign Images 1–4 only as sequential visual anchors. TIMELINE: 0–3s search and rack focus; 3–10s two fast scan transitions; 10–15s settle on the final mark. CONTINUITY LOCKS: the twin circular mask, vignette softness and installation lettering never change geometry. CAMERA: handheld micro-shake inside the mask only; cut at peak motion blur. AUDIO: wind, cloth movement and two short optical whooshes; restrained low pulse, no dominant score. END FRAME: hold the final mark sharp for one second. AVOID: mask drift, melted text, cross-dissolves and extra subjects.”

View full prompt

Use Images 1–4 as sequential keyframes, seen through a vintage binocular viewfinder searching for the MINIMAX installation. Open out of focus with subtle handheld shake, then push in quickly and rack focus onto Image 1. Between keyframes, use fast binocular-scan transitions with whip movement, motion blur, optical smearing, and brief exposure flicker. Cut at peak blur, then settle and snap back into focus. Keep the twin circular lens mask absolutely fixed throughout: identical position, scale, feathered black vignette, and edge softness, with no warping or drift. Only the image inside the mask may move. In Image 2, let the fabric move gently in the wind while the MINIMAX lettering follows the folds and remains legible. In Image 3, the subject should feel like a stylish passerby caught by chance, walking, turning, and swinging their arms naturally. In Image 4, the subject adjusts their glasses or lifts their chin slightly with a cool, effortless fashion-campaign attitude. Red typography should resolve with the focus: begin slightly blurred and at low opacity, then fade into clarity over 0.3–0.5 seconds. A subtle vertical slide or slight tracking expansion is allowed. Fade it out before the next transition or let motion blur carry it away. No spins, bounces, or large fly-ins/outs. Visual language: a voyeuristic, Wes Anderson-inspired 35 mm film look with fine grain, soft highlight halation, restrained color, and red typographic accents. Minimal, premium, lightly playful. Do not add people, vehicles, buildings, or logos. Preserve the core composition and the MINIMAX installation exactly. PRODUCTION CONTROL MODE: Reference to Video. Assign Images 1–4 only as sequential visual anchors. TIMELINE: 0–3s search and rack focus; 3–10s two fast scan transitions; 10–15s settle on the final mark. CONTINUITY LOCKS: the twin circular mask, vignette softness and installation lettering never change geometry. CAMERA: handheld micro-shake inside the mask only; cut at peak motion blur. AUDIO: wind, cloth movement and two short optical whooshes; restrained low pulse, no dominant score. END FRAME: hold the final mark sharp for one second. AVOID: mask drift, melted text, cross-dissolves and extra subjects.

Epic Space Opera Teaser AI video previewCinematic

Epic Space Opera Teaser

Cinematic
Model
MiniMax H3
Duration
5 seconds
Resolution
768p
Aspect ratio
16:9

“Epic theatrical space-opera teaser Keep the pace fast and the scale enormous without letting the edit drag. Use sharp hard cuts, a shaking command deck, white-hot flashes, split-second black frames, and a violent jump-to-warp impact. Title cards should use wide-tracked cinematic typography—not pure white—with restrained material texture, subtle illumination, and a faint edge glow. Animate the titles by emerging from deep-space shadow, catching a sweep of starlight, opening their letter spacing, leaving a slight afterimage, and flashing briefly against black. PRODUCTION CONTROL MODE: First & Last Frame. SHOT BUDGET: maximum four shots. TIMELINE: 0–3s command-deck disturbance; 3–7s gateway reveal; 7–11s jump-to-warp impact; 11–15s title on black. CAMERA: one primary movement per shot. TEXT LOCK: exact title copy, centered, wide-tracked, fully readable for at least 24 frames. AUDIO: sub-bass rise, one metallic impact at the warp cut, then a short vacuum drop. AVOID: extra title cards, soft dissolves, pure-white typography and constant camera shake.”

View full prompt

Epic theatrical space-opera teaser Keep the pace fast and the scale enormous without letting the edit drag. Use sharp hard cuts, a shaking command deck, white-hot flashes, split-second black frames, and a violent jump-to-warp impact. Title cards should use wide-tracked cinematic typography—not pure white—with restrained material texture, subtle illumination, and a faint edge glow. Animate the titles by emerging from deep-space shadow, catching a sweep of starlight, opening their letter spacing, leaving a slight afterimage, and flashing briefly against black. PRODUCTION CONTROL MODE: First & Last Frame. SHOT BUDGET: maximum four shots. TIMELINE: 0–3s command-deck disturbance; 3–7s gateway reveal; 7–11s jump-to-warp impact; 11–15s title on black. CAMERA: one primary movement per shot. TEXT LOCK: exact title copy, centered, wide-tracked, fully readable for at least 24 frames. AUDIO: sub-bass rise, one metallic impact at the warp cut, then a short vacuum drop. AVOID: extra title cards, soft dissolves, pure-white typography and constant camera shake.

Desert Fashion Campaign AI video previewFashion

Desert Fashion Campaign

Fashion
Model
MiniMax H3
Duration
5 seconds
Resolution
768p
Aspect ratio
16:9

“Create a premium 16:9 landscape fashion film. Use Image 1 for the overall mood, location, and film texture; Image 2 for the talent; Image 3 for the bag; and Image 4 for the closing brand mark. This is a fashion campaign for the clothing and bag. The tone is elevated, cool, and restrained, but the edit should still feel lively and fashion-forward—not like a conventional narrative film or an e-commerce ad. Keep the story simple: beside a vintage car on a desert highway, a woman walks to the rear of the car, opens the trunk, takes out a black bag, shares a quiet beat with the man standing nearby, then leaves carrying the bag. Integrate the clothing and bag naturally into the performance so they feel like part of the characters’ identity. PRODUCTION CONTROL MODE: Reference to Video. Image 1 controls desert location and film texture; Image 2 the woman; Image 3 the black bag; Image 4 the closing mark. ACTION ORDER: approach trunk → open trunk → lift bag by its handle → brief eye contact → walk away. CAMERA: three restrained editorial setups, never more than one move per setup. PRODUCT LOCK: preserve bag silhouette, material, hardware and color. AUDIO: desert wind, trunk latch, footsteps on gravel; minimal music. END FRAME: clean brand mark after the walk-away. AVOID: extra products, wardrobe swaps and e-commerce turntable shots.”

View full prompt

Create a premium 16:9 landscape fashion film. Use Image 1 for the overall mood, location, and film texture; Image 2 for the talent; Image 3 for the bag; and Image 4 for the closing brand mark. This is a fashion campaign for the clothing and bag. The tone is elevated, cool, and restrained, but the edit should still feel lively and fashion-forward—not like a conventional narrative film or an e-commerce ad. Keep the story simple: beside a vintage car on a desert highway, a woman walks to the rear of the car, opens the trunk, takes out a black bag, shares a quiet beat with the man standing nearby, then leaves carrying the bag. Integrate the clothing and bag naturally into the performance so they feel like part of the characters’ identity. PRODUCTION CONTROL MODE: Reference to Video. Image 1 controls desert location and film texture; Image 2 the woman; Image 3 the black bag; Image 4 the closing mark. ACTION ORDER: approach trunk → open trunk → lift bag by its handle → brief eye contact → walk away. CAMERA: three restrained editorial setups, never more than one move per setup. PRODUCT LOCK: preserve bag silhouette, material, hardware and color. AUDIO: desert wind, trunk latch, footsteps on gravel; minimal music. END FRAME: clean brand mark after the walk-away. AVOID: extra products, wardrobe swaps and e-commerce turntable shots.

Cyber-Grunge Fashion Film AI video previewCinematic

Cyber-Grunge Fashion Film

Cinematic
Model
MiniMax H3
Duration
5 seconds
Resolution
768p
Aspect ratio
16:9

“Use Image 1 as the reference for texture and mood, and Image 2 for the subject’s appearance. Generate a 15-second, 16:9 fashion short. Preserve the subject’s identity: long platinum-blonde hair, narrow black vintage sunglasses, a glossy black patent-leather trench coat, a cool, self-assured expression, and orange firelight reflected across the coat. Style: fast-cut fashion film on analog stock, set against a nighttime blaze with black smoke and orange-red flames. Layer in VHS glitches, CCTV signal interruptions, 1990s film grain, scanlines, chromatic aberration, light leaks, flash-to-white transitions, and subtle frame jitter. PRODUCTION CONTROL MODE: Reference to Video. Image 1 controls texture and fire-lit mood; Image 2 controls identity, hair, sunglasses and coat. TIMELINE: 0–4s close-up identity lock; 4–10s two fashion poses with hard cuts; 10–15s walking hero shot. EFFECT RULE: glitches may affect signal and background, not facial geometry or garment construction. CAMERA: handheld 35mm language with two deliberate flash cuts. AUDIO: fire crackle, low industrial beat and one tape-stop accent. AVOID: face replacement, coat color changes, continuous glitch overlays and unreadable silhouettes.”

View full prompt

Use Image 1 as the reference for texture and mood, and Image 2 for the subject’s appearance. Generate a 15-second, 16:9 fashion short. Preserve the subject’s identity: long platinum-blonde hair, narrow black vintage sunglasses, a glossy black patent-leather trench coat, a cool, self-assured expression, and orange firelight reflected across the coat. Style: fast-cut fashion film on analog stock, set against a nighttime blaze with black smoke and orange-red flames. Layer in VHS glitches, CCTV signal interruptions, 1990s film grain, scanlines, chromatic aberration, light leaks, flash-to-white transitions, and subtle frame jitter. PRODUCTION CONTROL MODE: Reference to Video. Image 1 controls texture and fire-lit mood; Image 2 controls identity, hair, sunglasses and coat. TIMELINE: 0–4s close-up identity lock; 4–10s two fashion poses with hard cuts; 10–15s walking hero shot. EFFECT RULE: glitches may affect signal and background, not facial geometry or garment construction. CAMERA: handheld 35mm language with two deliberate flash cuts. AUDIO: fire crackle, low industrial beat and one tape-stop accent. AVOID: face replacement, coat color changes, continuous glitch overlays and unreadable silhouettes.

Retro Anime Crime Title Sequence AI video previewCinematic

Retro Anime Crime Title Sequence

Cinematic
Model
MiniMax H3
Duration
5 seconds
Resolution
768p
Aspect ratio
16:9

“Create a 15-second, 16:9 opening-title sequence for a stylish crime mystery. Draw from retro Japanese animation titles, hard-edged silhouettes, comic-book collage, asymmetric split screens, bold geometric color fields, English credit typography, sparse katakana accents, and a noir-jazz sensibility. Aim for 60% suspense and 40% jazz: mysterious, cool, agile, and urban—never horror, heavy drama, or a cheerful jazz music video. Build the sequence as motion-graphics collage: linework appears over black, split-screen boundaries snap into place, then color blocks and frames assemble piece by piece. Character silhouettes, prop close-ups, and English credits slide, pop, and reveal through masks on the beat. Keep it designed and graphic, not like conventional narrative animation. Credits must be clean and legible. Animate thin frames drawing on, names sliding in, letters appearing one at a time, or color blocks revealing type, then hold briefly. Do not introduce Chinese text, garbled characters, or misspellings. Each role and each name appears once only; do not assign multiple roles to the same name. Vary the transitions: circular vinyl-record wipes, vertical car-door cuts, long character shadows wiping frame, red-line slices, oversized letter masks, split-screen reconfigurations, hard color-block cuts, and tiled frames snapping into place. Keep every transition crisp, rhythmic, suspenseful, and collage-driven. No soft dissolves or fluid morphs. BGM: an original 15-second title cue, 60% suspense and 40% jazz. Use a low drone, tense string pizzicato, cool synth pulses, low kick, sparse brushwork, fragments of walking bass, short baritone-sax phrases, and clipped brass accents. Build suspense with low frequencies and hi-hat in the first 2 seconds; bring in the low beat at 3 seconds, jazz-bass movement at 6 seconds, and a short sax/brass riff at 10 seconds. Freeze the final 2 seconds on a tense chord and drum hit. Keep it mysterious and criminally cool, never upbeat, and do not imitate an existing melody. PRODUCTION CONTROL MODE: Reference to Video. Images 1–5 define graphic language, character silhouettes, palette and credit layout; do not combine identities across references. SHOT PLAN: black-and-white linework → geometric split-screen assembly → prop close-ups → final ensemble title. TYPOGRAPHY LOCK: exact English credit strings supplied in the prompt, one credit per card, minimum one-second hold. MOTION: masks, slides and snap cuts on beat; no fluid morphing. AUDIO: brushed drums, upright bass and clipped brass accents. AVOID: narrative lip sync, 3D rendering, horror imagery and decorative fake text.”

View full prompt

Create a 15-second, 16:9 opening-title sequence for a stylish crime mystery. Draw from retro Japanese animation titles, hard-edged silhouettes, comic-book collage, asymmetric split screens, bold geometric color fields, English credit typography, sparse katakana accents, and a noir-jazz sensibility. Aim for 60% suspense and 40% jazz: mysterious, cool, agile, and urban—never horror, heavy drama, or a cheerful jazz music video. Build the sequence as motion-graphics collage: linework appears over black, split-screen boundaries snap into place, then color blocks and frames assemble piece by piece. Character silhouettes, prop close-ups, and English credits slide, pop, and reveal through masks on the beat. Keep it designed and graphic, not like conventional narrative animation. Credits must be clean and legible. Animate thin frames drawing on, names sliding in, letters appearing one at a time, or color blocks revealing type, then hold briefly. Do not introduce Chinese text, garbled characters, or misspellings. Each role and each name appears once only; do not assign multiple roles to the same name. Vary the transitions: circular vinyl-record wipes, vertical car-door cuts, long character shadows wiping frame, red-line slices, oversized letter masks, split-screen reconfigurations, hard color-block cuts, and tiled frames snapping into place. Keep every transition crisp, rhythmic, suspenseful, and collage-driven. No soft dissolves or fluid morphs. BGM: an original 15-second title cue, 60% suspense and 40% jazz. Use a low drone, tense string pizzicato, cool synth pulses, low kick, sparse brushwork, fragments of walking bass, short baritone-sax phrases, and clipped brass accents. Build suspense with low frequencies and hi-hat in the first 2 seconds; bring in the low beat at 3 seconds, jazz-bass movement at 6 seconds, and a short sax/brass riff at 10 seconds. Freeze the final 2 seconds on a tense chord and drum hit. Keep it mysterious and criminally cool, never upbeat, and do not imitate an existing melody. PRODUCTION CONTROL MODE: Reference to Video. Images 1–5 define graphic language, character silhouettes, palette and credit layout; do not combine identities across references. SHOT PLAN: black-and-white linework → geometric split-screen assembly → prop close-ups → final ensemble title. TYPOGRAPHY LOCK: exact English credit strings supplied in the prompt, one credit per card, minimum one-second hold. MOTION: masks, slides and snap cuts on beat; no fluid morphing. AUDIO: brushed drums, upright bass and clipped brass accents. AVOID: narrative lip sync, 3D rendering, horror imagery and decorative fake text.

Hand-Drawn Kitchen Creature AI video previewCinematic

Hand-Drawn Kitchen Creature

Cinematic
Model
MiniMax H3
Duration
5 seconds
Resolution
768p
Aspect ratio
16:9

“15 seconds, 16:9 landscape. Blend live-action footage of a small kitchen at dusk with hand-drawn luminous animation. The last sunset light lingers at the window. The lived-in kitchen contains an old wooden table, a half-washed mug, a lightly fogged glass bottle, and a hanging dish towel. Shoot as if someone is filming one-handed on a phone: subtle hand tremor, hesitant close-focus pulls, backlit exposure breathing, and slightly coarse noise in the shadows. It should feel like an astonishing event captured in a rush at home, not a carefully dressed commercial. Do not show giant eyes, split mouths, fangs, threatening behavior, lunges, sudden black frames, or jump scares. Use only room tone, cloth friction, a soft mug clink, faucet drips, the camera operator’s footsteps and quiet breathing, plus gentle electronic tones and tiny vocalizations from the drawn creatures. PRODUCTION CONTROL MODE: Text to Video. TIMELINE: 0–4s hesitant kitchen discovery; 4–9s luminous creature crosses the table; 9–13s it reacts to a mug clink; 13–15s operator steadies the phone. CAMERA: one-handed phone, delayed autofocus and exposure breathing. SOUND: room tone, faucet drip, cloth friction, mug clink, footsteps and breathing; only tiny electronic creature tones, no score. CONTINUITY: kitchen layout and sunset direction remain stable. AVOID: giant eyes, fangs, lunges, jump cuts to black and studio lighting.”

View full prompt

15 seconds, 16:9 landscape. Blend live-action footage of a small kitchen at dusk with hand-drawn luminous animation. The last sunset light lingers at the window. The lived-in kitchen contains an old wooden table, a half-washed mug, a lightly fogged glass bottle, and a hanging dish towel. Shoot as if someone is filming one-handed on a phone: subtle hand tremor, hesitant close-focus pulls, backlit exposure breathing, and slightly coarse noise in the shadows. It should feel like an astonishing event captured in a rush at home, not a carefully dressed commercial. Do not show giant eyes, split mouths, fangs, threatening behavior, lunges, sudden black frames, or jump scares. Use only room tone, cloth friction, a soft mug clink, faucet drips, the camera operator’s footsteps and quiet breathing, plus gentle electronic tones and tiny vocalizations from the drawn creatures. PRODUCTION CONTROL MODE: Text to Video. TIMELINE: 0–4s hesitant kitchen discovery; 4–9s luminous creature crosses the table; 9–13s it reacts to a mug clink; 13–15s operator steadies the phone. CAMERA: one-handed phone, delayed autofocus and exposure breathing. SOUND: room tone, faucet drip, cloth friction, mug clink, footsteps and breathing; only tiny electronic creature tones, no score. CONTINUITY: kitchen layout and sunset direction remain stable. AVOID: giant eyes, fangs, lunges, jump cuts to black and studio lighting.

MiniMax H3 workflow hub

Create, Learn, and Compare Outputs

Text to Video

Turn a written scene, action, camera direction, and sound brief into an AI video.

Open Text to Video

Image to Video

Animate a source image while directing motion, camera behavior, and scene changes.

Open Image to Video

Prompts

Copy practical MiniMax H3 prompt patterns, then send them to the generator.

Open Prompts

Generated Examples

Inspect MiniMax H3 generated videos with prompts, settings, and output descriptions.

Open Generated Examples

ComfyUI

Explore local MiniMax H3 workflows and ComfyUI setup guidance.

Open ComfyUI