The Higgsfield Prompt Vault: Every Workflow I Use, One Page

My attached photo is the identity reference. Generate 4 professional headshots of the person in the reference photo, preserving their exact likeness, facial features, skin tone, and hairstyle. One look per image: 1) CORPORATE: charcoal suit, crisp white shirt, soft light-gray studio backdrop. 2) STARTUP: plain black tee, moody dark backdrop, soft teal rim light. 3) CREATIVE: smart casual layers, warm tones, blurred golden-hour city background. 4) EDITORIAL: black and white, high contrast, dramatic side light. Every image: 85mm portrait lens look, f2.8 shallow depth of field, tack-sharp focus on the eyes, natural skin texture with pores (no plastic AI smoothing), relaxed confident expression with a slight natural smile, head-and-shoulders framing with a little headroom, 4:5 portrait crop, high resolution. No text, no watermarks, no logos, no accessories that are not in the reference photo.

MY CHARACTER: [swap this for any character you want, or keep mine: a small round fluffy cyber-fox with a soft plush midnight-blue and charcoal fur body, faint glowing teal circuit-line patterns woven through the fur, oversized expressive glowing cyan eyes, a small dark nose, two visible tiny front teeth, a cute surprised expression with a slightly open mouth, tiny paws, and a big fluffy tail with a softly glowing teal tip]

MY WORLD: [swap this for any setting you want, or keep mine: a magical bioluminescent cyber-forest at night, towering dark tree trunks with faint circuit-etched glowing bark, hanging fiber-optic vines with tiny cyan lights, glowing teal mushrooms on mossy ground, drifting light particles, gentle mist, volumetric light rays, cool teal rim light with a warm soft key light]

PROMPT 1, THE SCENE (image, 16:9, highest resolution):
Generate a cinematic 3D animated scene. My character stands in my world, large in frame and perfectly centered, full body visible, body facing forward. Composition rules, follow exactly: the head is slightly turned to the LEFT and the eyes are clearly looking toward the LEFT side of the frame. This exact starting pose is what makes the cursor tracking work later. Style: high-quality cinematic 3D animation, Pixar-like emotional charm, soft realistic detail, expressive character design, shallow depth of field, no text, no watermarks, no extra characters.

PROMPT 2, THE LEVEL PASS (animate the Prompt 1 image, 5 seconds, 1080p, silent):
The character slowly and smoothly turns its head from left to right at a steady natural pace, its eyes following the movement naturally, as if watching something glide across the screen in front of it. The body stays mostly still. Keep the expression, lighting, environment, and composition unchanged. No camera movement, no zoom, no cuts, no scene change.

PROMPT 3, THE DOWN PASS (first edit the Prompt 1 image so the head is tilted down with the eyes looking down-left at the ground, then animate that edited image):
The lowered head slowly turns from left to right, the gaze staying low the whole time, like it is watching something small move along the ground. The head never lifts. Same rules, no camera movement.

Run this pipeline on Higgsfield in strict order, one generation at a time. My attached photo is the identity reference for the main character. Do not skip steps, do not reorder, and wait for each generation to fully complete before using it as input to the next step.

STEP 1, MAIN CHARACTER SHEET (Nano Banana Pro, 3:4, my photo as reference image):
"Anime character sheet of the man in the reference photo, redrawn as a sports-anime football (soccer) protagonist in premium 2D anime style (crisp lineart, cel shading, dramatic rim lighting). Preserve his exact likeness: short dark curly hair with textured top and faded sides, full dark beard and mustache, dark brown eyes, medium-tan skin, strong straight eyebrows, athletic muscular build. Character sheet layout on a clean neutral studio background: full-body front view standing confidently, full-body three-quarter action pose dribbling a ball, and a close-up face portrait with intense determined eyes. He wears a completely fictional football kit: deep midnight-teal jersey with molten-orange diagonal lightning slash across the chest, orange trim on sleeves and collar, matching teal shorts and orange-teal socks, number 10 on the shorts, and an original invented crest of a stylized golden phoenix-wing shield (no real club, no real brand, no sponsor logos, no text or lettering anywhere on the kit). Consistent character design across all views, model-sheet quality, high detail."

STEP 2, SIDEKICK CHARACTER SHEET (Nano Banana Pro, 3:4, the STEP 1 result as reference image):
"Anime character sheet of a NEW second character in the exact same art style, lineart, cel shading and fictional football kit as the reference sheet: the protagonist's best-friend sidekick winger. Same team kit as the reference (midnight-teal jersey with molten-orange lightning slash, orange trim, golden phoenix-wing shield crest, no text or logos), but a clearly DIFFERENT person: lean fast winger build, longer straight jet-black hair swept back with an undercut, clean-shaven with light stubble, sharper jawline, playful confident grin, one eyebrow slash scar, medium-tan skin, dark eyes. Character sheet layout on clean neutral studio background: full-body front view, full-body sprinting action pose in the rain, close-up face portrait mid-shout. Consistent design across views, model-sheet quality, high detail, no lettering anywhere."

STEP 3, FINAL VIDEO (Seedance 2.0, 16:9, 12 seconds, BOTH sheets from steps 1 and 2 attached as image references; if a preset is suggested, decline it and generate literally):
"Cinematic 16:9 widescreen 2D sports-anime sequence, theatrical sakuga quality, three distinct shots, 12 seconds total. TWO consistent characters matching the two reference sheets exactly, both in the same fictional kit: midnight-teal jersey with molten-orange lightning slash, orange trim, golden phoenix-wing crest, no text or logos. CHARACTER A (protagonist): athletic man, short dark curly hair with faded sides, full dark beard, dark brown eyes, medium-tan skin. CHARACTER B (sidekick winger): lean build, longer straight jet-black hair swept back with undercut, clean-shaven with light stubble, eyebrow scar. Setting: packed stadium at night, cup final, torrential rain, blazing floodlights, wet glistening grass. SHOT 1, THE ASSIST: one continuous wide tracking shot, camera skimming low to the grass, racing alongside CHARACTER B as he sprints down the right wing through sheets of rain, boots kicking up spray, then cuts the ball back hard across the wet turf toward CHARACTER A arriving at the edge of the box, mouth open screaming to him, stadium bowl and floodlights filling the wide frame. SHOT 2, THE STRIKE, MAXIMUM IMPACT: extreme slow motion, low-angle close-up at grass level, CHARACTER A's planted boot slams down and the turf CRATERS, a crown of water and torn grass blasting outward, then his striking leg whips through the ball with a DETONATION: a single frame of white impact flash, a visible circular shockwave ring ripping outward through the rain, screen shake, the ball instantly IGNITES into a roaring fireball wrapped in molten-orange flame aura, and the camera whips violently to follow it as it streaks in a straight blazing line, a comet trail of fire and steam boiling the raindrops it passes. SHOT 3, THE GOAL AND POWER-UP: camera locked low behind the goal looking out through the net, mesh sharp in the extreme foreground, the fireball screams straight at the lens, goalkeeper diving full-stretch across frame and beaten, the burning ball rips into the TOP CORNER, the net BULGES violently toward the camera in a burst of fire, steam and spray, floodlights flaring into starbursts; then hard cut to CHARACTER A alone at the edge of the box POWERING UP: he plants his feet, clenches his fists and roars skyward as a towering column of molten-orange aura erupts around his body, lightning arcing off his shoulders, rain flash-vaporizing into steam around him, grass and water rippling outward in a radial shockwave, while the camera performs a fast dramatic 360-degree orbit around him, crowd and floodlights streaking past in the background, ember particles swirling. Cel-shaded anime, dramatic impact frames, speed lines, high-contrast rim lighting, film grain. Absolutely no on-screen text, captions, lettering, scoreboard numbers or watermarks."

Check your STEP 1 character sheet before running steps 2 and 3. Everything must be fully original: no real players, clubs, logos, brands, or existing anime characters.

Generate a viral man-on-the-street interview as a set of vertical 9:16 clips I can edit together.

THE HOST (the same woman in every clip):
An attractive woman in her mid 20s, styled and camera ready, holding a handheld microphone with a small blank mic flag. Warm, confident, playful energy. Busy city sidewalk in daytime, real pedestrians and street noise behind her.

THE LOOK:
Handheld camera, natural daylight, a slight push in on the punchy reactions, candid and slightly imperfect framing so it reads as real footage, not an ad.

THE QUESTION she asks every person:
"[YOUR QUESTION HERE]"

THE CAST (4 people, each a totally different look, age, ethnicity, and personality):
1. A confident young guy, quick and funny.
2. A stylish woman, a little sassy.
3. An older man, dry and matter of fact.
4. The plant: a relaxed, likeable person who casually answers "[PRODUCT OR PERSON YOU ARE PROMOTING], [one natural line on why it is great]" like it is obvious.

RULES:
Every answer is one or two short lines, spoken naturally, never a scripted ad tone. Build from mundane answers to the standout final one. Clear dialogue over ambient street audio.

OUTPUT:
One 3 to 5 second vertical clip per person (the host question plus their answer), plus one short clip of the host intro line. Delivered in order so I can stitch them straight together.

Here is a satellite screenshot with a red flight path line drawn on it. Identify the location and the real landmarks along the route. Then write a comprehensive FPV drone video prompt for an AI video generator: 9:16 vertical, photorealistic, 8 to 10 seconds. Start the video ON this map image, dive into it, and resolve into real aerial footage that flies the exact red line. Include a timed shot structure (0-1.5s, 1.5-4s, 4-6s, 6-8s) that names the actual landmarks the line passes, concrete altitudes and banking moves, and a style block: aggressive cinematic FPV, camera roll in turns, speed ramps, motion blur, golden hour light, no text, no watermarks, no propellers visible.

A cinematic, high-contrast motivational sequence in THREE shots, output in both 16:9 and 9:16. Pure black background throughout, hard single-source overhead lighting, deep falloff, moody film grain, shallow depth of field, 4K. Sports-documentary-meets-chiaroscuro tone — boxing-ring-at-midnight lighting. Two characters locked to reference images: YOUNG SELF (the boy, Image 1) and OLDER SELF (the muscular man in black tank top, Image 2). Identity must match the references exactly.

SHOT 1 (0–5s) — Tight straight-on frontal close-up of YOUNG SELF. Head and shoulders fill the frame, camera dead-center (not from the side), one hard light directly overhead against total darkness. He breathes heavily, chest rising and falling, sweat beads catching the light on his cheekbones and brow, eyes slowly lift further upward, lips part with a shaky exhale. Camera holds nearly locked with an extremely subtle, barely-perceptible forward drift — almost no zoom. Emotion carries the shot.

SHOT 2 (5–8s) — Over-the-shoulder from BEHIND YOUNG SELF. His back-of-head and shoulder silhouette fill the left foreground, soft rim light on his hair. Across a pitch-black void stands OLDER SELF in his own narrow overhead beam, mid-distance, clearly visible from torso up, fists clenched, facing camera. Slow continuous dolly push-in past the young self's shoulder toward the older self — OLDER SELF grows slightly larger as we approach. Dust particles drift through his beam. Tense, still.

SHOT 3 (8–13s) — Extreme wide profile two-shot on a BIG dramatic sloped mound rising from lower-left to upper-right. The ground reads as polished BLACK OBSIDIAN — glassy, near-mirror dark with faint specular highlights where light touches it. YOUNG SELF stands far LEFT at the lower base, in profile facing right. OLDER SELF stands far RIGHT at the elevated top, in profile facing left, clearly higher — significant horizontal separation between them. Each figure is hit by a sharp, high-definition VOLUMETRIC PILLAR of light coming STRAIGHT DOWN from directly above their head — perfectly vertical god-light columns, never angled. Volumetric haze visible only inside the beams. Everything outside the beams is crushed deep black — no ambient spill, no rim, the void between the pillars reads as absolute darkness. Camera executes a smooth, substantial horizontal dolly-back at locked eye-level elevation — NO upward tilt, NO crane, horizon line locked. By the end the figures shrink visibly, revealing the full sweep of the obsidian slope and the towering vertical beams. Subjects hold still in profile, mythic stillness.

Style: epic, emotional, mythic, restrained. Hard cuts between shots, no transitions. Final delivery upscaled to 4K.

PROMPT 1, THE HERO VIDEO (Seedance 2.0, 9:16, 8 seconds):
Cinematic hero video for [BRAND: two sentences on what it is and who it is for]. Open on a moody close-up detail shot, pull back to reveal the product or subject in its environment, end on a clean centered hold with shallow depth of field, premium color grade matched to the brand's vibe, soft volumetric light. No text, no watermarks, no logos.

PROMPT 2, THE AD IMAGES (Nano Banana Pro, 4:5, run three times, one style line each):
High-end advertising photo of [BRAND]. Style: [run 1: clean studio, seamless background, controlled soft light / run 2: candid lifestyle scene, real environment, natural light / run 3: bold neon night look, hard rim light, high contrast]. Sharp focus, premium detail, no text, no watermarks.

PROMPT 3, THE B-ROLL PAIR (Seedance 2.0, 9:16, 5 seconds, run twice):
Atmospheric b-roll clip that matches the vibe of [BRAND]: [run 1: a slow macro detail shot with drifting light / run 2: a smooth gliding movement through the brand's environment]. Cinematic grade, shallow depth of field, no text, no people speaking to camera.

PROMPT 4, THE THUMBNAILS (Nano Banana Pro, 4:5, run three times):
Bold thumbnail-grade image for [BRAND], strong single subject, dramatic lighting, high contrast against a deep dark background, instantly readable at small size. No text, no watermarks.

PROMPT 5, THE BANNER (Nano Banana Pro, 16:9):
Wide banner image for [BRAND], subject placed off-center with clean negative space on one side for a headline to be added later, premium cinematic grade consistent with the rest of the kit. No text, no watermarks.

Take this [SPORT] clip of [PLAYER or TEAM]. Cut it into a vertical 9:16 funk-style hype edit, 15 to 20 seconds, timed to a hard Brazilian funk / phonk beat. Sync every cut and zoom-punch to the beat hits. Find the biggest moment, the [GOAL], and hit it with a slow-motion speed ramp on the bass drop. Dark moody cinematic grade with heavy contrast, a glow and aura on [PLAYER], light leaks, motion blur, and a camera shake on the drops. Bold glitchy text with their name. Make it a premium funk sports edit.

Take this [SPORT] highlight: [paste the YouTube link]. Cut it into vertical 9:16 clips, 15 to 20 seconds each, focused on the biggest moments, the goals and the skills. Export them ready to post.

Generate a 2D animated kids cartoon clip (Seedance 2.0, 16:9, 10 to 12 seconds), wholesome and original, for ages 3 to 6. MY CHARACTER (keep identical every episode): [a cheerful round baby blue elephant with big friendly eyes, tiny yellow rain boots, and a red scarf]. THIS EPISODE'S MICRO-LESSON: [washing your hands is fun]. The character acts out the lesson in one simple scene with a clear beginning and happy ending, bright saturated colors, soft rounded shapes, gentle bouncy motion, simple uncluttered background. A cheerful sing-along melody carries the clip. No text, no watermarks, no scary elements, no brand characters.

skincare drops, glowing skin, hands applying serum

30-second Reel spec, every video:
- Footage: Higgsfield (free credits) dark gym or sunrise, or royalty-free B-roll
- Voiceover: ElevenLabs free voice, deep tone, reading one quote
- On screen: the quote, nothing else
- Length: 30 seconds

Take the top 10 sellers, run Meta ads on only those. Higgsfield generates the UGC unboxing video.

My attached photo is the product reference, preserve its exact shape, color, materials, and any real labels. Generate 6 high end ecommerce product shots of this exact product, one look per image: 1) pure white seamless studio background, soft even light, clean catalog style. 2) on a natural surface (light wood or stone) with a soft window light shadow. 3) dramatic dark background, single hard rim light, premium hero shot. 4) floating on a solid pastel background with a soft contact shadow, modern DTC look. 5) lifestyle scene, the product in the real environment it is used in, shallow depth of field. 6) top down flat lay with a few tasteful complementary props. Every image: sharp focus, realistic materials and reflections, true to life color, 4:5 crop, high resolution. No added text, no watermarks, no logos that are not on the real product.

My attached photo is the pet reference, preserve the animal's exact breed, fur color, markings, and face. Paint this pet as a noble 17th century Dutch royal oil portrait: the pet wearing an ornate period costume (velvet doublet, lace collar, a small jeweled medallion), seated in a regal three quarter pose against a dark chiaroscuro background, warm candlelit key light, visible oil brushwork and canvas texture, an implied gilded frame. Museum quality classical painting, dignified expression, no text, no watermarks. 4:5 portrait.

My attached photo is the reference for the body part and skin. Realistically render this tattoo onto the exact area shown, following the skin's contour, curvature, and lighting so it looks genuinely inked and healed, not a flat sticker: [DESCRIBE THE TATTOO, e.g. a fine line black and grey snake wrapping the forearm, roughly 15cm]. Match the ink saturation and slight skin translucency of a real tattoo, preserve the person's exact skin tone, hair, and the original lighting and pose. Photorealistic, no text, no watermarks.

My attached photo is the room reference. Redesign this exact room in a [STYLE, e.g. warm Japandi / mid century modern / dark moody industrial] style. Keep the room's real architecture identical: same wall positions, window placement, ceiling height, and camera angle. Only change the furniture, decor, textiles, lighting, and color palette to fully match the style. Photorealistic interior render, natural daylight from the existing windows, realistic materials and shadows, magazine quality staging, no text, no watermarks.

My attached photo is an empty room in a real property. Virtually stage it: add tasteful, realistic furniture and decor appropriate for a [ROOM TYPE, living room / bedroom / office] in a [TIER, mid range / luxury] home. Keep the room's real walls, floor, windows, and camera angle exactly as shown. Only add furnishings that fit the real proportions and cast correct shadows in the existing light. Clean, aspirational, buyer ready. Photorealistic, no text, no watermarks.

My attached photo is the identity reference, preserve the person's exact face and likeness. Reimagine this person authentically in the [ERA, e.g. 1920s / 1970s / 1880s Old West / far future cyberpunk]. Period accurate hairstyle, wardrobe, and styling, shot on the photography of that era (correct film stock, color or sepia or black and white, lens character, grain, and framing). Change only the era styling, keep the real facial features exactly. Photorealistic period portrait, no modern elements, no text, no watermarks. 4:5.

My attached photo is the identity reference for the hero. Create one dynamic comic book page (American superhero style, bold inked lineart, halftone shading, vivid colors) starring this exact person as the protagonist, preserving their real likeness. Lay out 3 to 4 panels telling this micro story: [DESCRIBE THE BEAT, e.g. they discover a power, face a threat, and win]. Dramatic angles, motion lines, dramatic lighting. Leave every speech balloon and caption box EMPTY (blank white) so I can add text later. No pre written text, no watermarks. Portrait page, high resolution.

My attached image is a brand logo, preserve it exactly (colors, shapes, proportions). Generate 4 photorealistic product mockups with this logo applied naturally and correctly, following the surface curvature, fabric weave, lighting, and perspective, never a flat paste: 1) embroidered on a folded premium hoodie. 2) printed on a matte ceramic mug on a cafe table. 3) etched on a storefront glass door. 4) debossed on a leather bound notebook. Realistic materials, studio grade lighting, no extra text beyond the logo, no watermarks. 4:5.

My attached photo is the reference for this exact dish, preserve the food's real ingredients, portion, and plating. Reshoot it as a premium food photography hero image: styled on a beautiful plate and surface that suits the cuisine, appetizing fresh garnish or gentle steam where natural, gorgeous soft directional light, shallow depth of field, rich appetizing color, a few tasteful props (utensils, ingredients, linen) around the edges. Mouth watering, editorial quality, true to the real dish, no text, no watermarks. 4:5.

Animate a charming brand mascot video (Seedance 2.0, 9:16, 6 to 8 seconds). Turn [SUBJECT, e.g. my product / my logo / a coffee cup] into a cute expressive character with friendly eyes and a mouth, keeping its real shape and colors recognizable. It looks directly at camera, gives a warm little wave or bounce, and delivers a short friendly greeting with natural mouth movement and expression. Clean brand colored background, soft studio light, playful Pixar like charm, subtle bounce animation. No on screen text, no watermarks.

Generate a set of vertical 9:16 close up food b-roll clips, 3 to 4 seconds each, cinematic and ASMR satisfying, for this recipe step: [STEP, e.g. cracking an egg into flour / searing a steak / drizzling honey]. Extreme macro, shallow depth of field, slow motion on the key action, glistening textures, natural light, gentle steam or motion. Appetizing, tactile, premium cooking show grade. No hands to camera talking, no text, no watermarks. Deliver as separate clips I can stitch in order.

Generate a set of cinematic travel b-roll clips (Seedance 2.0, 9:16, 5 to 8 seconds each) for [DESTINATION, e.g. Kyoto in autumn / the Amalfi Coast at golden hour]. Deliver a variety pack: one sweeping aerial establishing shot, one slow ground level street or nature detail, one iconic landmark reveal, one atmospheric golden hour mood shot. Photorealistic, rich color grade, natural motion, travel documentary quality, no recognizable faces in focus, no text, no watermarks.

My attached photo is the exact vehicle reference, preserve its model, color, wheels, and details. Create a cinematic 9:16 automotive commercial (Seedance 2.0, 8 seconds): the car driving through [SETTING, e.g. a rain slick neon city at night / a golden hour desert highway / a misty coastal mountain road], dramatic reflective light sweeping across the bodywork, low tracking and orbiting hero angles, spray or dust kicking up, glossy premium color grade. End on a clean slow motion hero hold. Photoreal, high end car ad energy, no text, no added logos, no watermarks.

Create a cinematic teaser trailer sequence (Seedance 2.0, 9:16, 10 to 12 seconds, several hard cut shots) for [CONCEPT, e.g. my app / my personal brand / a fictional film about X]. Epic movie trailer energy: a moody establishing shot, a rising tension middle beat, and a punchy hero climax shot, dramatic lighting, film grain, anamorphic lens flares, big cinematic color grade, pacing built around a beat to drop on. Leave every frame text free so I can add title cards later. No watermarks.

Create a satisfying before and after transformation clip (Seedance 2.0, 9:16, 6 seconds). Start on the BEFORE state: [DESCRIBE BEFORE, or attach it as the reference], then smoothly morph or wipe into the AFTER state: [DESCRIBE AFTER] across the clip so the change feels dramatic and rewarding. Match lighting and framing between the two so it reads as the same subject transformed. Clean, high contrast, scroll stopping, no text, no watermarks.

Generate a short horror atmosphere clip (Seedance 2.0, 9:16, 6 to 8 seconds) for a fictional scary story narration channel. Scene: [DESCRIBE, e.g. an empty hospital corridor with one flickering light / a dark forest treeline at night]. Slow creeping camera push, deep shadows, cold desaturated grade, subtle fog, a single unsettling detail (a distant figure, a slowly opening door), heavy film grain, tension without a cheap jump scare. Cinematic dread, no text, no watermarks.

Generate a perfectly seamless looping 9:16 clip (Seedance 2.0, 6 seconds) of an oddly satisfying process: [PICK ONE, kinetic sand being sliced / paint colors swirling and blending / a machine folding something precisely / glossy spheres cascading]. Macro focus, buttery slow motion, hyper clean studio lighting, rich saturated colors, and a first frame that matches the last frame exactly so it loops with no visible cut. Satisfying, tactile, no text, no watermarks.

Generate a striking 1:1 album cover for a [GENRE, e.g. dark synthwave / lo-fi / drill / dream pop] track. Mood: [DESCRIBE, e.g. a neon lit lonely highway at night / a surreal dreamlike garden]. Bold, memorable, gallery grade art direction that fits the genre, a strong focal subject, cinematic color, and clean space left for a title to be added later. No pre written text, no watermarks. High resolution square.

My attached photo is the exact garment reference, preserve its real cut, color, pattern, and details. Generate 4 editorial fashion lookbook photos of a photorealistic model wearing this exact garment, styled cohesively: varied natural poses, a [SETTING, e.g. minimalist studio / sunlit city street / concrete gallery] backdrop, soft directional fashion lighting, premium magazine editorial grade, realistic fabric drape and texture. Keep the garment faithful to the reference. Believable model, no text, no watermarks, no brand logos that are not on the garment. 4:5.

Create a punchy 9:16 social ad (Seedance 2.0, 8 seconds) for a local [BUSINESS TYPE, e.g. barbershop / coffee shop / gym]. My attached photo is the real location or product, keep it recognizable. Open on an energetic hook shot, show the space or service looking its absolute best with warm inviting light and smooth camera motion, and end on a clean hero hold with space for a logo and offer to be added later. Vibrant, local premium feel, no text, no watermarks.

Create a cinematic real estate tour clip (Seedance 2.0, 16:9 or 9:16, 8 seconds) from my attached room photo, keeping the room's real layout and finishes. Smooth gliding gimbal style camera motion moving forward through the space and revealing it, bright natural window light, a warm inviting grade, a calm premium listing feel, subtle depth. Photorealistic, faithful to the real room, no text, no watermarks.

Children's picture book illustration (Nano Banana Pro, 4:3, warm storybook style, soft painterly texture). MY CHARACTER (keep identical on every page): [e.g. a small brave hedgehog named Pip with a patched red scarf, round glasses, and a tiny satchel]. THIS PAGE'S SCENE: [describe the moment and setting]. Wholesome, gentle, richly colored, one clear focal action, plenty of soft negative space at the top for a line of text to be added later. Consistent character design, no pre written text, no watermarks.

My attached photo is the identity reference, preserve the person's exact face. Generate 4 portraits in the style of cheesy 1990s American high school yearbook and mall glamour shot photography: soft focus dreamy lighting, laser or mottled studio backdrops, period accurate hair and wardrobe, slightly faded film color, the exact posed vibe of the era. One look each: classic senior portrait, soft glow beauty shot, moody laser background, dramatic black and white. Keep the real face exactly. Nostalgic and believable, no text, no watermarks. 4:5.

Generate a clean 9:16 background video (Seedance 2.0, 8 seconds, subtle looping motion) for a faceless [THEME, e.g. flags / geography / movie] trivia channel. A visually rich themed scene with gentle continuous motion (drifting particles, slow parallax) and a calm uncluttered center where a question and options will be overlaid later. Vibrant but not distracting, premium mobile game polish, no text anywhere, no watermarks.

My attached photo is the identity reference, preserve the person's exact face and outfit if visible. Realistically place this person into [SCENE, e.g. standing on a Santorini clifftop at sunset / on safari beside a jeep]. Match the scene's lighting direction, color temperature, perspective, and shadows so the person is convincingly part of it, with correct scale and ground contact. Photorealistic travel snapshot look, candid framing, no text, no watermarks.

Generate a photorealistic UGC style ad video (Seedance 2.0, 9:16, 8 seconds). A relatable everyday person, real and believable rather than a polished model, filming themselves selfie style in a natural home setting, holding [MY PRODUCT, attach a photo as the reference and keep it exact], talking directly to camera with warm genuine energy as if recommending it to a friend. Handheld phone camera look, natural window light, casual and authentic, subtle natural mouth movement matched to a friendly delivery. The product stays clearly visible and true to the reference the whole time. No on screen text, no watermarks. Deliver a hook shot and a product in hand close up so I can cut them together.

Generate a bright 16:9 animated kids YouTube scene (Seedance 2.0, 10 to 15 seconds, run several scenes and stitch them into one episode), wholesome and original, for ages 2 to 6. MY RECURRING CHARACTER (keep identical in every single video): [e.g. a happy little yellow robot named Bloop with big round eyes, a glowing light up tummy screen, and squeaky wheels for feet]. THIS EPISODE teaches: [MICRO LESSON, e.g. counting to five / the colors of fruit / the ABCs]. The character cheerfully demonstrates the lesson step by step in a simple colorful world, big friendly expressions, gentle bouncy motion, a repeating sing along melody, and a slow clear pace a toddler can follow. Bright saturated colors, soft rounded shapes, uncluttered background, no scary elements, no existing brand characters, no text, no watermarks.

Generate cinematic open world game style clips (Seedance 2.0, 9:16, 6 to 8 seconds each, run several) for a faceless gaming channel, in the look of a modern AAA crime open world game cinematic. Everything must be original, no real game, brand, logo, or existing character. MY CHARACTER: [DESCRIBE, e.g. a hooded street hustler in a worn leather jacket]. Deliver a variety pack: a slow cinematic character reveal on a rain slick neon city street, a fast low chase shot driving through downtown at night, a moody rooftop shot overlooking the skyline, and a tense first person walk down an alley. Photoreal leaning game cinematic grade, dramatic lighting, filmic depth of field, completely HUD free and text free, no real logos, no watermarks.

Generate a seamless looping website hero background video (Seedance 2.0, 16:9, 8 seconds, subtle slow motion). Subject: [DESCRIBE THE VIBE, e.g. slow drifting abstract liquid gradients in my brand colors / a calm aerial glide over clouds at dawn / soft floating particles in a dark studio]. Very slow, calm, continuous motion that never pulls focus, and a first frame that matches the last frame exactly for a clean invisible loop. Leave generous flat, low contrast negative space so headline text stays readable on top, with the darker calmer area toward the [left / center] for the copy. Premium, ambient, no text, no watermarks.