Most AI video models stop at 10-15 seconds. Seedance 2.5 renders up to 30 seconds natively — with its own audio track — in a single pass, at 480p or 720p.
Real generations from our video models — Veo 3.1, Kling and Seedance. Hover any clip to play, click the speaker icon to hear the audio.
Hover to play · Click speaker to unmute
Real generations from our video models — Veo 3.1, Kling and Seedance. Hover any clip to play, click the speaker icon to hear the audio.
Alien Fantasy World
Luxury Perfume Ad
Product Launch Teaser
Maldives Travel Commercial
Eastern Wuxia Duel
Hollywood Racing Movie
Spaceship to the Stars
Golden Retriever & Butterflies
Sushi Chef at Work
Red Dress on the Cliffs
Midnight Duel
Handheld Selfie Vlog
Mouth-watering Food Close-up
Urban Street Dance
Luxury Lipstick Ad
One native pass, picture and sound together, up to half a minute long. No stitching, no continuation calls, no drift between segments.
Pick 5, 10, 15, 20, 25 or 30 seconds. The whole clip is rendered in one native pass, so characters, lighting and pacing stay consistent from the first frame to the last.
Seedance 2.5 writes the soundtrack while it renders the picture. Ambience, effects and motion-matched sound arrive with the video instead of in a separate pass.
Give it a starting image, and optionally an ending image, and Seedance 2.5 fills in the motion between them. Ideal for product reveals, transformations and before-and-after arcs.
16:9, 9:16, 1:1, 3:4, 4:3 and 21:9. Landscape ads, vertical shorts, square feed posts and widescreen openers all come out of the same model.
In text-to-video, attach up to 9 reference images to hold style and character steady, and up to 3 reference audio tracks to steer the sound. Optional web search pulls in real-world context.
480p costs 8 credits per second, 720p costs 18. A 5-second draft starts at 40 credits, a full 30-second 720p cut is 540, and failed generations are refunded in full.
Seedance 2.5 handles picture and sound in the same pass. You describe the whole clip once — there is nothing to stitch together afterwards.
Write the entire duration as one continuous description: setting, subject, movement and sound. Choose your length, aspect ratio and resolution before generating.
The model plans motion, lighting and pacing across the whole duration, then renders the video and its matching audio track together in a single pass.
Take the finished video with audio baked in — no watermark, commercial rights included. If a generation fails, your credits come straight back.
Longer runtime with sound changes what you can actually finish in one generation — full spots and complete story beats, not just a single moving shot.
A 30-second spot is the standard broadcast and pre-roll length. Setup, demonstration, benefit and sign-off fit inside one generation, sound included.
TikTok, Reels and Shorts in 9:16, from a 5-second hook to a 30-second story. Audio comes attached, so nothing needs a soundtrack pass afterwards.
Half a minute is enough for a real narrative arc: establish the situation, show the change, land the point — with continuous characters throughout.
Aerial sweeps, coastal drives and street-level wandering that need time to breathe — plus the wind, water and city sound that sells the place.
Thirty seconds is a lot of screen time to fill. These habits keep Seedance 2.5 on script from the first frame to the last — and keep the job from being misread.
Describe the finished clip, don't hand it instructions
Verbs like edit, extend, continue, delete or replace can make the request read as a video-editing job rather than a generation job, and the task fails. Describe what the finished clip looks like instead.
Example
"Avoid: "extend this clip and replace the sky". Prefer: "the same harbour at dusk under an orange sky with low cloud, the boat moving slowly from left to right across the frame"."
A 30-second clip needs a beginning, a middle and an end, or the model will hold one idea for half a minute. Write the progression in the order it should happen.
Example
"The room starts dim and empty, morning light climbs the far wall, dust drifts through the beam, and the camera slowly settles on a cup of coffee going cold."
Audio generation is on by default, so anything you leave unsaid gets invented for you. One or two lines of sound direction is usually enough to keep it in character.
Example
"Sound: steady rain on a tin roof, an occasional low roll of thunder in the distance, no music."
Stunning, epic and cinematic tell the model nothing. Materials, colours, light direction and strong verbs do.
Example
"Avoid: "a stunning epic mountain shot". Prefer: "granite ridges under low side light, snow spilling off the crest in a thin plume"."
Seedance 2.5 offers 5, 10, 15, 20, 25 and 30 seconds, six aspect ratios and two resolutions. Deciding before you generate saves credits and keeps the composition right.
Example
"A 15-second 9:16 vertical clip: a barista pulling an espresso shot, steam rising into warm café light, shallow depth of field."
The two modes are separate. Image-to-video takes a first frame, and optionally a last frame, and fills in the motion between them. Text-to-video instead accepts up to 9 reference images and up to 3 reference audio tracks.
Example
"First frame: a sealed cardboard box on a table. Last frame: the box open with a camera inside. Prompt: "hands lift the flaps in one smooth motion, soft window light"."