Flux 3 Video vs Seedance 2

Vertical 9:16 "3 ways to style this one piece" fashion UGC video for Instagram Reels, ultra-realistic and shot on an iPhone camera. Total duration: 12 seconds.

Use @Image1 as the character reference. Maintain the same face, hairstyle, body type, skin tone, and overall identity throughout the full video. Character persona: graceful ethnic-wear creator — poised, warm, and traditional in movement.

This format shows the SAME core garment styled in different ways. The hero piece must be visually consistent across every beat — only the accessories, footwear, layering, or styling combinations change between cuts.

Use @Image2 as Way 1, @Image3 as Way 2, and @Image4 as Way 3 — each is the same core piece styled differently. Preserve the core garment exactly across all 3 beats; vary only the accessories, footwear, layering, or styling cues that distinguish each way.

Use @Image5 as the setting reference: a warm Chandni Chowk-style boutique corner — soft warm light from the right, wooden display shelf and brass accents visible, patterned rug on the floor. Keep the setting CONSISTENT across all 3 ways.

Camera: vertical iPhone footage, mostly static, full-body or 3/4-body framing, slight natural handheld feel. SAME camera position and framing across all 3 beats (framing: chest-to-toe, centered, distance unchanged).

Timing: three equal beats, ~4.0s each. Use hard jump cuts between beats.

[0–4.0s] Beat 1 — Way 1
- Action: Creator walks in from left and stops in center, holds one confident pose showing full outfit. Minimal movement (one hand adjusting dupatta).
- Styling: use @Image2 styling exactly. Preservation: preserve the core garment's color (rich maroon embroidered kurta), longline flowy silhouette, and visible dupatta fabric and border — do not alter embroidery, length, or color.

[4.0–8.0s] Beat 2 — Way 2
- Hard jump cut. Action: same framing; creator does a 3/4 turn to camera and places hand on hip to show changed styling.
- Styling: use @Image3 styling exactly. Preservation: preserve the core garment's color (rich maroon embroidered kurta), longline flowy silhouette, and embroidery details — only change accessories/layering as in @Image3 (e.g., add belt or tuck) and footwear; do not alter fabric or pattern.

[8.0–12.0s] Beat 3 — Way 3
- Hard jump cut. Action: same framing; creator faces camera and gives a final confident pose with slight shoulder pop.
- Styling: use @Image4 styling exactly. Preservation: preserve the core garment's color (rich maroon embroidered kurta), longline flowy silhouette, and embroidery details — only add outer layer or jacket and accessories as shown in @Image4; do not alter the garment itself.

Style direction: comparative styling reel, warm festive vibe, natural skin tones, soft golden highlights from right-side light. Keep movements graceful and deliberate; avoid chaotic camera moves or extreme zooms.

VO (spoken, Hinglish) — Dialogue mode ON. Use this exact script as the only spoken dialogue / voice-over, delivered naturally by the creator and timed across the three beats. Do not add extra spoken CTAs.
- Exact script (must be used verbatim): "3 ways to wear this—Anmol Collections. Dupatta loose, belt karke, jacket add; har look easy aur festive. Link in bio, check karo."
- VO timing suggestion: deliver the first clause in [0–2s] as the hook, middle styling callouts across [2–9s], and final CTA line in [9–12s]. If physically cannot fit, trim only minimally but do NOT rewrite brand wording or language.

Music: brisk upbeat snippet with a clean beat-drop at each jump cut (at 4.0s and 8.0s).

Restrictions: no on-screen text overlays, no added people, no clothing morphs — use hard cuts only. Preserve identity and all facial features from @Image1 throughout.

Output-ready notes for production team: use @Image1 for the actor; style each beat to match @Image2, @Image3, @Image4 respectively; use @Image5 as the set dressing reference; keep each beat ~4s, VO exact Hinglish script as above.

@Audio1 is the complete and only spoken performance for this video. The output's voice track must be this exact audio, word for word — same words, same pacing, same pronunciation. If a person is on camera while speech plays, their lips sync to @Audio1 verbatim. Do not generate, add, or substitute any other speech, narration, or voice-over. Background music stays subtle under the voice.

Vertical 9:16 mirror-selfie fashion UGC video for TikTok / Instagram Reels, ultra-realistic and shot on an iPhone camera. Total duration: 12 seconds.

Use @Image1 as the character reference. Maintain the same face, hairstyle, body type, skin tone, and overall identity throughout the full video. Persona: graceful Brand-clinic ethnic creator — poised, warm, traditional energy.

Use @Image2 as the outfit reference. Preserve the exact outfit design, color, sharara silhouette, hand-embroidery detail, fabric drape, slight stretch fit, and visible embellishment from @Image2. Preservation note: keep the sharara set's color, flowy flared silhouette, and hand embroidery detail exactly as shown; do not alter fit, pattern, or embellishment placement.

Use @Image3 as the mirror / dressing-area setting reference. Setting description: a Chandni Chowk-style dressing area with a full-length mirror, warm ambient light from the right, tiled or wooden floor and a visible dressing table surface — realistic reflection of the room as in @Image3.

Camera: mirror-selfie style iPhone capture. The creator (from @Image1) holds the phone naturally while filming herself in the mirror. Use realistic reflection framing, slight natural handheld movement, and clear outfit visibility. No on-screen text overlays or brand text rendered into the video.

Action progression (three beats, timed to 12s total):

[0–4s] Beat 1 — Stand and turn: Creator stands in front of the mirror wearing @Image2, holding the phone as in a mirror-selfie. Small body turn to show the full sharara silhouette; steady framing so the full outfit is visible in the mirror. Spoken dialogue (voice-over / natural lip-sync): "Dekho yeh sharara set. Hand embroidery, halka stretch fit, bahut comfortable. Anmol Collections. Link se check karo." (Begin speaking immediately; natural paced delivery fits within the 12s total.)

[4–8s] Beat 2 — Reframe and side pose: Creator angles the phone slightly to capture the outfit better, shifts weight to a side pose to show flared sharara movement and embroidery close to camera. Maintain soft confident expression and gentle handheld motion. Continue same spoken line if still mid-sentence, natural cadence.

[8–12s] Beat 3 — Final outfit-check: Creator steps slightly closer or back for a clear hero hold of the outfit in the mirror, pauses with a subtle head tilt and soft smile for a clean final shot. End with the last words of the supplied script. Hold final frame cleanly for the last second.

Style: authentic mirror-check content, platform-native, casual yet traditional. Music: soft lo-fi or chill Indian pop bed at low volume under dialogue; no aggressive drops. Avoid distorted reflections, identity or outfit drift, extra limbs, chaotic camera movement, and any inserted on-screen brand text.

VO placement: full supplied Hinglish script must be the only spoken dialogue/voice-over. Trim only if physically cannot fit; do not rewrite wording.

Deliverable: single continuous 12s mirror-selfie clip shot as described, using @Image1, @Image2, @Image3 references and preserving the outfit details exactly.

@Audio1 is the complete and only spoken performance for this video. The output's voice track must be this exact audio, word for word — same words, same pacing, same pronunciation. If a person is on camera while speech plays, their lips sync to @Audio1 verbatim. Do not generate, add, or substitute any other speech, narration, or voice-over. Background music stays subtle under the voice.

Vertical 9:16 fashion fit-review UGC video for TikTok / Instagram Reels, ultra-realistic and shot on an iPhone camera. Total duration: 12 seconds.

Use @Image1 as the character reference. Maintain the same face, hairstyle, body type, skin tone, and overall identity throughout the full video.

Use @Image2 as the outfit reference. Preserve the exact outfit design, color, silhouette, fabric feel, print, pattern, embroidery, fit, and major visible details from @Image2.

Use @Image3 as the setting reference. Describe setting exactly as in @Image3: a warm indoor Chandni Chowk–style boutique corner with soft overhead warm light, wooden floor, and a decorative mirror or traditional textile backdrop visible — natural practical lighting from the right side.

Camera: vertical iPhone footage, creator talking-to-camera style, natural UGC framing with a mix of medium and full-body shots. Keep the outfit clearly visible.

Action progression (three beats, scaled to 12 seconds total):

[0–4s] Beat 1 — Intro reaction
- Medium shot, creator facing camera wearing the outfit from @Image2, natural honest smile and slight head-nod. She lightly adjusts the neckline or dupatta once with a slow, natural hand movement to show fabric feel. VO (same timing): spoken line in Hinglish begins immediately.

[4–8s] Beat 2 — Demonstrate fit
- Full-body to 3/4 turn: creator turns slightly to right then left to show front and side fit; she takes two small walking steps forward to demonstrate how the outfit falls and moves. Quick close-up of embroidery when she pauses with hand near the bodice.

[8–12s] Beat 3 — Final verdict
- Medium shot, creator gives a satisfied confident pose (hand on hip or gentle dupatta toss), smiling to camera for the final look. Hold final hero frame as VO finishes.

Dialogue / Voice-over (Hinglish, lip-sync on-screen): use ONLY this exact line — do not add or change language: "Anmol Collections — yeh suit perfect fit hai, soft fabric aur detailed embroidery. Ghar baithe order karo, link pe jao." Align VO timing across the three beats so speech is clearly audible and sits above minimal ambient bed.

Style: honest creator review, natural and trustworthy, realistic fashion UGC, warm indoor light, clear outfit visibility.

Sound: minimal — soft ambient room tone with a barely-there acoustic bed; speech sits on top.

Preservation instructions:
- Outfit (@Image2): preserve the exact color, longline ethnic salwar silhouette, flowy fabric drape, and the visible detailed embroidery on the bodice and borders as shown in @Image2; do not alter fit or embellishment placement.

Technical notes / constraints: hard cuts only between beats; no on-screen text overlays or added brand text; single person in frame only; avoid chaotic camera moves or aggressive zooms; natural hand movements only.

@Audio1 is the complete and only spoken performance for this video. The output's voice track must be this exact audio, word for word — same words, same pacing, same pronunciation. If a person is on camera while speech plays, their lips sync to @Audio1 verbatim. Do not generate, add, or substitute any other speech, narration, or voice-over. Background music stays subtle under the voice.

Vertical 9:16 festive get-ready fashion UGC video for Instagram Reels, ultra-realistic and shot on an iPhone camera. Total duration: 12 seconds.

Character persona: graceful ethnic-wear creator — warm, poised, familiar with traditional dressing. Use @Image1 as the character reference. Maintain the same face, hairstyle, body type, skin tone, and overall identity throughout the full video.

Outfits & jewellery preservation:
- Preserve @Image2 exactly: the visible garment is a festive ethnic sharara-style outfit — keep the color, flared silhouette, hand-embroidered embellishment, lightweight fabric drape, and any dupatta or sequin/embroidered details exactly as seen.
- Preserve @Image3 exactly as the jewellery set: keep the design, metal tone, stonework, shape (jhumkas/necklace/bangles if visible) and placement exactly as shown.

Setting: use @Image4 as the getting-ready space reference. Describe the space exactly as in @Image4: a cozy dressing corner with warm ambient light, a wooden surface (vanity or low table) and a visible mirror or soft textured backdrop. Light direction: warm golden-side fill; surfaces: wooden/soft fabric; visual anchors: the mirror and a small tray for jewellery.

Camera: vertical iPhone handheld feel. Medium-framing for ritual moments, 3/4 or full-body framing for final reveal. Use hard jump cuts between outfit stages (no morphing).

Dialogue / voice-over: Dialogue mode ON. Use this exact spoken script in Hinglish, delivered naturally by the on-screen character as voice-over or lip-synced speech: "Diwali ready? Anmol Collections ka hand-embroidered sharara, lightweight comfort, perfect for family puja. Shop the collection." Do not add any other spoken CTA or text. Ensure timing fits the 12s cut.

Beat structure and precise timing (total 12s):
[0–3s] Beat 1 — Casual base, holding the festive piece: Medium shot of @Image1 in a simple cotton base layer near the vanity from @Image4, holding or smoothing the sharara from @Image2 laid on the surface. Small anticipatory smile, gentle fingertips on embroidery. No jewellery yet. Natural warm side light.

[3–7s] Beat 2 — Hard cut to wearing the festive outfit: Jump cut. @Image1 now wearing the exact outfit from @Image2. 3/4 framing as she adjusts the dupatta or smooths the sharara flare with one deliberate motion. Keep movement slow and natural; preserve outfit details and drape.

[7–10s] Beat 3 — Jewellery layer: Close-medium shot as she puts on the jewellery from @Image3 (necklace or jhumkas visible). Show the action of clasping/placing one piece; avoid extreme hand close-ups. Preserve the jewellery design exactly.

[10–12s] Beat 4 — Final festive reveal: Full-body or mirror-facing shot. She turns to camera or mirror, gives a confident soft smile and holds a poised final pose showing the full look (outfit + jewellery). Warm golden framing.

VO beat map and sync: Fit the supplied Hinglish line across the full 12s with natural pacing (approximately 2 words/sec). Start spoken hook in Beat 1 and complete CTA by Beat 4. Ensure lip-sync or voice-over matches the cuts; do not shorten or rewrite wording except only if physically impossible to fit.

Style & mood: warm festive get-ready, sensorial and intimate, realistic UGC. Music: warm modern-traditional bed (soft sitar or subtle tabla) low under VO. Avoid on-screen text, logos, or rendered brand text. Avoid multiple people in frame and any clothing morphing transitions.

Final notes: keep continuity of the same person (@Image1) across all cuts, keep all outfit and jewellery details identical to @Image2 and @Image3, and use the exact setting from @Image4.

@Audio1 is the complete and only spoken performance for this video. The output's voice track must be this exact audio, word for word — same words, same pacing, same pronunciation. If a person is on camera while speech plays, their lips sync to @Audio1 verbatim. Do not generate, add, or substitute any other speech, narration, or voice-over. Background music stays subtle under the voice.

Vertical 9:16 fabric and detail close-up reel for Instagram Reels, ultra-realistic and shot on an iPhone camera. Total duration: 12 seconds.

Character persona: graceful ethnic-wear creator (soft, traditional, poised). Use @Image1 only as an off-camera presence: a wrist/hand entering frame to touch or lift fabric; match skin tone from @Image1. Never show the face.

Setting: use @Image3 as the surface/backdrop reference — a warm textured backdrop with soft directional warm side-light and a low-reflective surface that lets the garment colors and embroidery pop. Keep surfaces matte and minimal; avoid extra props.

Garment: use @Image2 as the garment hero. Preserve the exact outfit design, color, silhouette, fabric texture, print/pattern, embroidery/beadwork, neckline finish, and visible craft details from @Image2.

Preservation instruction for outfit (@Image2): Preserve the exact color, longline ethnic silhouette, and the resham + zardozi embroidery placement and stitch texture visible on @Image2 — do not alter thread color, motif density, or embellishment finish.

Format notes: product-first, slow and sensorial. Mostly still camera with very slow deliberate pushes/pans. One slow cut at most. No face frames, no jump cuts, no on-screen text overlays, no brand text rendered in-frame. Minimal ambient acoustic bed only.

VO (Hinglish, spoken exactly as written, reverent slow tone; total pacing ~1.5 words/sec): "Chandni Chowk ki zardozi aur resham kaam, aaraamdayak fit, shandar kapda. Anmol Collections. Dekho aur mangwao." Do not add or change words. Sync VO across beats as indicated below.

Camera/action progression and timing (total 12s):
[0–4s] Beat 1 — Wide hero: static or very-slow push-in wide shot of the full garment laid on or draped over the @Image3 surface so the silhouette, color, and overall embroidery layout read clearly. Off-camera hand from @Image1 may gently smooth a fold at frame edge. VO: first phrase "Chandni Chowk ki zardozi aur resham kaam," timed to introduce craft.

[4–8s] Beat 2 — Detail close-up: slow smooth iPhone pan to a tight close-up on a primary craft detail from @Image2 (embroidered neckline or prominent motif). Hold steady, reveal stitch texture and metallic sheen. No shake. Off-camera finger may lightly trace fabric edge (no blocking of embroidery). VO: middle phrase "aaraamdayak fit, shandar kapda."

[8–12s] Beat 3 — Texture & finish: another slow close-up on a secondary detail or the drape — gently lift a corner with the off-camera hand to show fluidity and inner finish, or show beadwork catching light. End on a calm hero hold of the most striking detail. VO final phrase "Anmol Collections. Dekho aur mangwao." Conclude with a soft fade-out in audio and image.

Additional constraints: no fast movements, no aggressive zooms, avoid harsh shadows or specular hotspots on metallic threads, keep motion minimal and reverent. Use only the supplied VO; no extra spoken CTAs.

@Audio1 is the complete and only spoken performance for this video. The output's voice track must be this exact audio, word for word — same words, same pacing, same pronunciation. If a person is on camera while speech plays, their lips sync to @Audio1 verbatim. Do not generate, add, or substitute any other speech, narration, or voice-over. Background music stays subtle under the voice.

Format: Vertical 9:16, 12 seconds. Use provided images as exact assets: @Image1 is the hero product/treatment photo; @Image2 is the real vanity/studio setting and must be used for the ritual scene. Preserve the product packaging: the product is a tube (or pump-style bottle if visible) in a muted pastel color with a cylindrical/rounded shape — keep label visible but not legible. Tone / voice: Affordable Gen-Z / problem-solving, real-skin, believable improvement. Dialogue mode ON — spoken audio only (Hinglish script exactly as provided). No on-screen text generated in-scene; any tags or brand text are post only. Natural sound cues (dropper click/lather/soft towel) layered under music. Camera: vertical iPhone UGC framing, close macro and short handheld moves. No full faces; focus on cheek/hand/neck area or product application area shown in @Image1. Hard cuts between beats, realistic skin texture, subtle glow in after shot. Timecodes and beats: [0–3s] Beat 1 — Dull (first ~25%): Start with a muted, slightly desaturated close-up pulled from the aesthetic of @Image1 showing the 'before' skin state (dull, slightly textured cheek or hand) in soft, flat light. Slow 1–2 second micro-pan to show honest texture. Ambient foam/drip sound muted. No talking yet. [3–8s] Beat 2 — The ritual (middle ~40%): Hard cut into the vanity from @Image2 — warm, slightly backlit. Show the hero using the product from @Image1: one deliberate application gesture (squeeze or pump then rub between fingertips, then gentle upward strokes on cheek/hand). Include two close macros: (a) product texture on fingertips, (b) application stroke. Preserve packaging appearance exactly as described. Diegetic sounds: product click/pump and soft rubbing; music opens slightly. Voice-over (start at ~3.2s, intimate tone, only this exact line): "Skin bejaan lag rahi hai? Roz face wash + SPF50 se nikhra aur protection milega. Kaleigh Beauty, ek baar try karo." (Use natural, friendly Hinglish; sync to the ritual so words flow over the application.) [8–12s] Beat 3 — Glow (final ~35%): Hard cut back to the original framing from Beat 1. Brighter warm grade, subtle specular highlights on hydrated skin, realistic pore texture visible. Single small gesture (brush hair back or light exhale/soft smile off-camera) to sell confidence. End on a calm hold: the product from @Image1 resting next to the glowing cheek/hand in the @Image2 vanity environment for the final 0.8–1s. Music swells gently then fades. Audio notes: spoken script is the only VO; no additional spoken copy or CTAs. Keep improvement believable (hydration, subtle glow), no airbrushing or impossible changes. Avoid on-screen text, legible labels, numbers, price/claims visually. Deliver as single 12s vertical cut with hard cuts between beats.

@Audio1 is the complete and only spoken performance for this video. The output's voice track must be this exact audio, word for word — same words, same pacing, same pronunciation. If a person is on camera while speech plays, their lips sync to @Audio1 verbatim. Do not generate, add, or substitute any other speech, narration, or voice-over. Background music stays subtle under the voice.

Concept: A 12s vertical GRWM, no-talking-on-camera UGC focused on a single hero product @Image1 with vanity setting from @Image2. Realistic Indian creator (young adult, fresh-faced skincare user) — hair clipped back, natural skin texture, subtle imperfections allowed. Preserve packaging: the product is shown as a white squeeze tube with a matte finish and rounded cap (keep label visible but not legible). Keep results believable: subtle brighter, hydrated skin.

Beats (strict timecodes):
[0–2.4s] Beat 1 — Sit-down: Creator enters frame at the vanity visible in @Image2, hair clipped back, sits facing the mirror/phone tripod. Static vertical framing, natural daylight or soft vanity light. No mouth movement. Natural diegetic sounds (mirror tap, seat shift). Spoken VO starts immediately: "Aaj ka look ready—salicylic-glycolic face wash plus SPF 50 gives skin protection. Kaleigh Beauty, try it." (VO only; face remains silent.)

[2.4–9.0s] Beat 2 — Steps run (jump-cut rhythm ~115–120 BPM): Quick jump-cut 1: small prep dab (splash of water or towel gesture) [2.4–3.3s]. Jump-cut 2 (longest close-up) [3.3–6.6s]: deliberate application with the white squeeze tube @Image1 — clear hand-to-face gesture, slow gentle lather/texture push-in macro, keep packaging visible in hand. Jump-cut 3 [6.6–7.8s]: blend/rinse-off or pat motion at the mirror, natural small adjustments (hair tuck). Keep framing consistent for snap cuts. VO continues over these cuts within the same delivered script line.

[9.0–12.0s] Beat 3 — The look (final): Hard cut to final mirror check — creator turns slightly to camera with a quiet confident half-smile (mouth closed), then looks to mirror. Place the white tube (@Image1) on the vanity (from @Image2) in-frame lower third. Hold final pose ~2.5–3s. End on natural ambient vanity sound; VO concludes within the 12s.

Camera & style notes: Vertical iPhone UGC framing, mix of medium vanity shots and a macro texture push-in for the product application. Lighting soft and flattering, no heavy beauty filters. Keep skin pores/texture visible. No on-screen text, no labels rendered legible, no additional products, no impossible skin transforms, and no model lip-syncing. Music: muted trending GRWM pop at low volume behind VO; add small diegetic taps and water sounds.

Audio requirement: Use only the supplied Hinglish spoken script exactly: "Aaj ka look ready—salicylic-glycolic face wash plus SPF 50 gives skin protection. Kaleigh Beauty, try it." Timing must fit within 12s; trim only if absolutely necessary.

@Audio1 is the complete and only spoken performance for this video. The output's voice track must be this exact audio, word for word — same words, same pacing, same pronunciation. If a person is on camera while speech plays, their lips sync to @Audio1 verbatim. Do not generate, add, or substitute any other speech, narration, or voice-over. Background music stays subtle under the voice.