Start from a traceable source case
Use the paired public X post and verbatim prompt as one reference record. The collection orders those records newest-first by X Post ID and does not rewrite, popularity-rank, or claim ownership of creator material.
SEELE TV / Prompt archive
Explore 181 receipt-validated Seedance 2.5 video prompts with more than 200 English words: 165 current-page records and 16 preserved historical-only records. Every card keeps the complete prompt, author, directly playable source video, self-hosted first-frame poster, lightweight original-post link, copy control, and SEELE Workspace handoff together.
One source post / one complete prompt
Verbatim prompts, click-to-load source video, and a lightweight link to each original X post.
Exact mapped collection
Prompt · 580 English words · current daily source card
CAMERA: Handheld DV 16mm daily vlog footage. The video MUST begin with Emma holding the camera at arm's length in selfie mode while stepping outside her apartment building with her dog. The entire video remains handheld selfie-style. The camera itself is never visible. Keep subtle handheld shake, drifting composition, gentle autofocus hunting, small rushed reframing, slight uneven movement, exposure breathing, occasional imperfect framing, and natural motion blur throughout. The footage should feel genuinely handheld and imperfect, like a real person's casual morning vlog. LOOK: Warm analog tape texture with gentle film grain, slightly softened sharpness, subtle halation around sunlight, realistic skin tones, low contrast, tiny exposure shifts, and natural motion blur. Authentic everyday footage rather than a polished commercial. STYLE: Relaxed morning lifestyle vlog. Quiet, cozy, spontaneous, and natural. Emma speaks casually to the camera with short, believable sentences and comfortable pauses. She occasionally looks at her dog while talking. Keep the pacing slow and realistic. Do NOT rush actions or dialogue. CHARACTER: EMMA — a beautiful white woman in her mid-20s. Long light brown hair in a messy ponytail, green eyes, minimal makeup, oversized gray hoodie, black biker shorts, white sneakers, and a small crossbody bag. She is walking a happy golden retriever. SETTING: A peaceful suburban neighborhood on a sunny morning. Tree-lined sidewalks, quiet residential streets, birds singing, a small neighborhood park with benches, green grass, and soft golden morning light. Very few people are around. SCENE — 30 SECONDS 0–8 SECONDS: The vlog begins as Emma steps outside her apartment building with her golden retriever on a leash. She holds the camera in selfie mode and smiles naturally. She says: "Good morning. I swear, this guy has been waiting by the door for like ten minutes." She looks down at the dog and smiles. "He's ready to go." Keep her movements natural and slow. The dog stays beside her. 8–16 SECONDS: Emma walks slowly down the quiet sidewalk with the dog. The dog stops to sniff a tree. Emma pauses and looks down at him. She laughs softly: "Of course we have to stop at every single tree." She waits for the dog instead of pulling him immediately. "Take your time." The dog finishes sniffing and they continue walking. 16–23 SECONDS: They enter the small neighborhood park. Emma continues holding the camera in selfie mode while walking slowly. She briefly turns the camera slightly toward the dog, then back toward herself. She says: "It's actually so nice out this morning." She looks around at the trees. "I really should come here more often." Keep the camera movement smooth and natural. Do not introduce any new actions. 23–30 SECONDS: Emma continues walking through the park with the dog beside her. She looks down at him and smiles. She says: "He's definitely happier than I am right now." She laughs quietly. After a short pause, she looks back into the camera: "Anyway, coffee time. See you later." She gives a small natural wave and ends the recording. IMPORTANT: The entire video is exactly 30 seconds long. Prioritize realistic pacing, continuity, and visual accuracy over fitting in more events. Do not compress actions. Do not make Emma move or speak unnaturally fast. Every action should happen sequentially and have enough time to render correctly. Keep the dog movement simple and physically realistic. No cuts. No editing. No montage. No time-lapse. No camera placement. No additional characters. No additional dialogue. No sudden scene changes. The entire video should feel like one continuous authentic morning vlog generated from a single prompt.
Prompt · 260 English words · historical snapshot
POV motocross sequence on an erupting volcano. The entire scene is from the rider’s perspective, with helmet-camera realism, natural head movement, violent suspension shake, dust on the lens, heat haze, flying ash and intense engine sound. The rider never drives on lava. He stays strictly on rocky volcanic paths, ash-covered trails, loose black gravel and narrow ridges while avoiding lava flows and glowing cracks. 0-5s: POV accelerating downhill on a steep black volcanic trail. Gloved hands, handlebars and front fender visible. Lava flows are visible nearby on both sides at a distance, but the bike remains on solid rocky ground. Small rocks bounce across the trail, smoke rises in the background. 5-10s: The rider weaves through falling debris and sharp turns on a narrow ash-covered path. A fresh lava stream cuts across the route ahead, so he sharply turns onto a higher rocky side trail, avoiding the lava completely. 10-15s: The path narrows along a ridge. The rider jumps over a glowing crack in the ground, clearing it cleanly without touching the molten area below, lands hard on solid volcanic rock, and keeps control while larger stones tumble behind him. 15-20s: A dense wave of ash rolls down the slope. The rider threads between two collapsing boulders, descends a final rocky path, and bursts onto a safer lower ridge far from the lava. He briefly looks back toward the erupting crater while still riding away at speed. Photorealistic motocross physics, realistic suspension movement, coherent volcanic terrain, practical disaster realism, no third-person shots, no fantasy look, stable bike geometry, intense POV immersion. #DreaminaCPP @dreamina_ai
Prompt · 507 English words · historical snapshot
SHOT 1: ECU, 85mm push-in / on the phone screen as it shakes on rumpled sheets / SFX: alarm, sheet rustle. SHOT 2: WS, 35mm handheld jolt / Rhythmic cut into her jolting upright through side light, throwing the blanket aside, planting her feet in one rushed motion / SFX: mattress bounce, blanket whip, sharp breath. SHOT 3: MCU, 50mm slide / Cut on action into face wash at the sink, droplets catching the top light / SFX: faucet rush, water slap. SHOT 4: Insert, 85mm lens focus / Match cut into the toothbrush, mint foam and mirror eye / SFX: bristle scrape, sink drip. SHOT 5: Interior fridge view, 24mm wide / Object pass into the camera inside the fridge looking out as the door snaps open, blue light framing a hurried grab / SFX: fridge hum, bottle clink, shelf scrape. SHOT 6: Insert, 50mm handheld / Rhythmic cut into egg and toast hitting the pan under warm practical light / SFX: butter sizzle, chop tap. SHOT 7: MCU, centered 50mm push-in / Match cut into one rushed bite, a clock glance, an immediate rise / SFX: crunch, ceramic clink, chair scrape. SHOT 8: Bird’s-eye, 35mm overhead / Match cut on action as the sleep tee disappears under a fitted top and jacket; keys, transit card, and bag scooped in one grab / SFX: fabric whip, key jingle, zipper pull. SHOT 9: MS, 35mm pan / Camera slips into the rushed outfit change / SFX: fabric whip, key jingle, bag rustle. SHOT 10: Insert, 50mm overhead / Match cut into lace-up shoes slamming on, laces yanked tight in one impatient pull / SFX: sole thump, lace tug, short breath. SHOT 11: WS, 24mm parallel / Whip pan transition into her rushing through the corridor into bright light without breaking stride / SFX: latch click, rapid footsteps, hallway air. SHOT 12: MS to CU, 35mm glide into 85mm push-in / Sound bridge into the metro as she grips the pole and slips through the closing doors, chrome streaking, city blurring / SFX: train rattle, door chime, light breath. SHOT 13: Insert to MCU, 50mm snap zoom / Smash cut to the office entrance, access card on the reader, glass door unlocking, dropping into her chair / SFX: badge beep, door click, laptop chime. SHOT 14: OTS, 35mm handheld / Rhythmic cut into fingers racing across the keyboard, blue monitor light on her face / SFX: keyboard clicks, notifications, mouse click. SHOT 15: WS, 24mm pull-out / Match cut back to the bedroom composition from the opening shot as she collapses onto the bed / SFX: body drop, fabric rustle, heavy exhale. A few things worth noticing about why it’s built this way: Every shot names its transition in. Not “then this happens” but cut on action, match cut, object pass, sound bridge. The connection is specified, so the sequence has rhythm instead of just order. Shot 15 deliberately matches Shot 2’s composition. Same bedroom, same framing, opposite energy. That bookend is what turns a montage into a day. The SFX line is separate on every shot and never repeats. Bristle scrape, shelf scrape, lace tug. Audio detail at that resolution is what makes fifteen fast cuts feel like one continuous space. And the lens changes carry meaning, not variety. 85mm for the intimate inserts, 24mm for the wides, a snap zoom for the smash cut. Focal length is doing narrative work.
Showing 3 of 3 matching cases; 3 total.
How to use the Seedance 2.5 archive
Use the paired public X post and verbatim prompt as one reference record. The collection orders those records newest-first by X Post ID and does not rewrite, popularity-rank, or claim ownership of creator material.
Name the identity, wardrobe, object, geography, screen direction, lighting, sound, and motion details that must survive the sequence. Remove adjectives that do not change a frame or an audible event.
Inspect physical behavior, shot transitions, subject consistency, audio, pacing, and delivery fit. Revise the smallest failed instruction before adding more cuts or visual effects.
Evidence boundary
Use the source, rights and current workspace controls as the acceptance boundary.
Before you hand off
The source snapshot contains 291 receipt-validated records. The canonical prompt-length gate publishes 181 and quarantines 110; prompt text is preserved verbatim, each published case links to its matching post and author, and first-frame posters are generated from the validated source video.
Yes. The complete direction and the Seedance 2.5 template identity are carried into the generation URL for review in the workspace.
Continue the workflow
Continue in the SEELE workspace to inspect the currently available Film & CG workflow.
Try it free