A model from ByteDance for cinematic video generation with native audio. The cheapest per-second price, 6 aspect ratios, First+Last Frame morphing.
Seedance 1.5 Pro is a model by ByteDance (the creators of TikTok).Generates cinematic video with native audio,with camera control and support for 6 aspect ratios including CinemaScope 21:9.The lowest per-second price on the platform.
💡 Useful tips:
1:1, 4:3, 3:4, 16:9, 9:16 and CinemaScope 21:9 — works for any content
Sound effects, background music and voices — synced to on-screen motion
From 2 cr/sec — cheaper than any other model on the platform
Generate the entire video from a text description — without images
Bring 1 image to life or morph between 2 (First+Last Frame)
Static camera for dialogue and portraits. Turn off for dynamic shots
Generate video from text — without uploading images, the model creates everything from scratch.
Animate 1 image or create a smooth transition between 2 frames (First+Last Frame).
Seedance uses a scripting approach — write the prompt as a scene description.Order the elements strictly — the model prioritizes by position in the text.
Prompt example:
"A street musician sits by a brick wall in daylight and plays an acoustic guitar. The camera pans slowly left to right, moving from his hands on the strings to his face. A passerby drops a coin into the open case. Sounds: rhythmic acoustic guitar, light city ambience, footsteps on the sidewalk. Soft, natural light, documentary aesthetic."
Describe camera motion in plain words: pan, tilt, zoom, rotation. Specify direction, speed, and trajectory.
The model links visual motion to sound. Contradictions yield unpredictable results.
"A quiet thunderstorm with strong wind and bright lightning" — 3 conflicting signals
"A distant thunderstorm with quiet thunder and a light wind" — one dominant tone
💡 Our prompt enhancement system helps assemble the description in the right format
The model already SEES everything in the image. Describe only what should to change.When uploading 2 images, describe how the transition between them happens.
Actions, camera motion, physics (wind, light), sounds, transitions between frames
Appearance, clothing, hair color, background details — the model takes them from the photo
Price = price per second × duration. Depends on resolution and whether audio is enabled.
Study the prompts for high-quality video — these examples help you grasp the structure of effective descriptions
Premium photorealistic fantasy realism. Medium shot: Subject raises hand, draws semicircle with blue magical glow and sparks forming glowing "Sixio" inscription. Cut to close-up: soft excited smile, gaze forward, inscription blurred in background. Audio: winter wind, magical bells, soft female voice "Sixio, can you hear me?" with English lip-sync, calm orchestral New Year melody. Volumetric lighting, high detail.
Beneath the red walls of the Forbidden City on a beautiful day, a chubby orange cat wearing sunglasses lounges arrogantly on a traditional armchair like a boss. It lazily speaks in a Beijing accent: "得嘞，今儿个天儿不错，来晒晒太阳".
Elegant woman in flowing evening gown walks away, turns head, reaches out invitingly and says "Shall we dance?" Elegant pose, luxury hotel corridor, natural body proportions, refined silhouette, soft golden hour lighting, fashion editorial style, shot from behind at 45 degree angle, cinematic composition, fashion magazine aesthetic.
An internal light source from the mech's chest shines outward as the mech passes through the hangar's light curtain and ascends directly into the sky, achieving a smooth transition from mechanical interior space to open air.
Soft light pours down, silk shimmering with a pearlescent sheen. She gently walks toward the camera, the skirt draping and flowing, outlining a form that is both soft and luxurious.
On an aerial corridor in a steampunk city, a mechanical technician stands on a suspended bridge, rapidly communicating with an assistant via radio in Cantonese: "气压表又掉了，帮我开三号阀！" The camera looks up from beneath the bridge, emphasizing the dizzying height. The massive sounds of gears and pistons blend with the wind, forming a complex environmental soundscape.
Text→Video for generation from scratch. Image→Video — upload 1 image (animation) or 2 (First+Last Frame morphing).
Pick a format (16:9, 9:16, 21:9...), resolution, duration, and audio. Fixed Lens — for a static camera.
Enter a description in any language (except terms like Dolly zoom, Tracking shot). AI will enhance the prompt with the proper element structure.
Press 'Translate to English' — AI will translate the prompt for free. Then press 'Generate'. Result in 2-5 minutes.
Seedance 1.5 Pro is a model from ByteDance (creators of TikTok) for generating cinematic video with synchronized audio. Supports text→video and image→video (including First+Last Frame).
480p (standard), 720p (high), and 1080p (Full HD). Pick 480p for drafts and tests — it's the cheapest option.
The model automatically generates sound effects, background music, and voices synchronized to on-screen motion. Audio is optional, enabled separately, and affects the price.
Fixed Lens locks the camera — the frame stays static, only the subject moves. Disable for dynamic flythroughs and pans. Useful for dialogue and talking heads.
Yes! 1 image — classic image-to-video. 2 images — First+Last Frame mode: the model creates a smooth transition between the start and end frames.
Seedance supports 6 aspect ratios: 1:1, 21:9 (CinemaScope), 4:3, 3:4, 16:9 (landscape), and 9:16 (portrait). The widest choice among our models.
Cost = price per second × duration. Per-second price depends on resolution (480p/720p/1080p) and audio. The cheapest option is 480p without audio.
Seedance is the best choice for: budget clips and ads, animating photos (image-to-video), morphing between two frames (First+Last), video with native audio, and content needing many aspect ratios (including 21:9 CinemaScope).
Cinematic video with native audio, 6 aspect ratios, First+Last Frame — from 2 cr/sec
We use cookies to operate the service, keep your session, and collect anonymous statistics. See our Privacy Policy.