Choose a category
3 models available
4 models available
3 models available
2 models available
4 models available
8 models available
Seedance 2.5 Text to Video
Write it out and get up to 30 seconds with sound.
Creative Vision
Setting, subject, action, camera. Dialogue in quotes gets spoken.
Video Parameters
Priced per second — 30 seconds costs 7.5× what 4 does.
720p costs 2.25× what 480p does. Draft at 480p.
Adaptive only makes sense with reference material — pick a ratio here.
AI audio synced to the video. Does not change the price.
Cinema Preview
Generation History0 records
Seedance 2.5 Text to Video
Write a shot and Seedance 2.5 films it — up to 30 seconds in a single run, at 480p or 720p, with audio generated alongside the picture rather than dubbed on afterwards. ByteDance's 2.5 generation is built for long takes, so this is the page to use when you want a scene rather than a five-second fragment.
Thirty seconds is a lot of prompt to write — find the shot at 480p and 4 seconds first
Why Choose Seedance 2.5 for Text to Video
Most models on this site cap out between 10 and 15 seconds. This one runs to 30, generates its own audio, and charges strictly by the second — so length is a decision you make, not a limit you hit.
Up to 30 Seconds in One Run
The slider runs 4 to 30 seconds, one second at a time. No stitching and no continuation trick — a single generation covers a whole beat: an entrance, a turn and a settle.
Audio Comes With It, Free
Sound is produced in the same pass as the picture, so dialogue lands on the mouth and effects land on the action. It is on by default and does not change the price — unlike Seedance 1.5 Pro, where audio doubles the cost.
Priced Per Second, Nothing Hidden
30 credits buys 4 seconds at 480p, 75 buys 5 seconds at 720p, 430 buys the full 30 seconds at 720p. There is no per-run surcharge and no long-clip penalty — the number scales with exactly what you asked for.
Seven Aspect Ratios
16:9, 9:16, 1:1, 4:3, 3:4, 21:9 and Adaptive. Vertical for social, 21:9 when the shot should read as cinema — decided before generation rather than by cropping afterwards.
Add a Reference Clip and Pay Less
Under Advanced you can attach a video of up to 30 seconds to lend its motion, or an audio track to drive the delivery. The reference video also drops the provider's rate by about 40%, and the price on the button follows.
Failed Runs Are Refunded
Credits are held when you submit and returned automatically if the generation fails or the provider never answers. On a model where one clip can cost 430 credits, that is not a footnote.
The Long-Take Model
If you need 4K, Seedance 2.0 goes there. If you need cheap, 1.5 Pro starts at 2 credits. This page is for when the shot needs room to breathe.
How to Use Seedance 2.5 Text to Video
Long clips fail differently from short ones: a five-second shot only has to look right, a thirty-second shot has to go somewhere. Write the movement, not just the picture.
Describe the Shot and the Move
Setting, subject, action, camera — then say how it develops across the clip. "Holds, then pushes in as the lights come up" gives the model 30 seconds of structure. Dialogue in quotation marks gets spoken. The box holds 20,000 characters.
Draft Short and Small
480p at 4 seconds is 30 credits. Run the prompt two or three ways at that setting to see how the model is reading your wording before you pay for length.
Then Stretch It
Once the wording is settled, push the duration up. Price is linear, so 20 seconds costs exactly five times 4 seconds — there is no threshold where it suddenly gets expensive.
Finish at 720p
720p costs 2.25 times what 480p does per second, so switch to it for the take you intend to keep. Generation takes a few minutes and you can leave the page — the result appears in your history.
Need to Start From a Picture?
Seedance 2.5 also runs from a still, from a start-and-end pair, or from a cast of reference images — same model, same price, different page.
Seedance 2.5 Text to Video
FAQ
Length, audio, pricing, aspect ratios and how this page relates to the other Seedance 2.5 modes.
Seedance is ByteDance's video generation family, and 2.5 is its long-form multimodal generation: up to 30 seconds per run with audio produced in the same pass. This page drives it from a prompt alone. FlowVeo3 is an independent platform and is not affiliated with ByteDance.
Between 30 and 430 credits, set by resolution and length: 30 credits for 4 seconds at 480p, 35 for 5 seconds, 75 for 5 seconds at 720p, and 430 for a full 30 seconds at 720p. The exact cost for your settings is shown before you submit.
Yes. The provider produces the whole clip in one pass rather than stitching shorter pieces, which is why the motion stays coherent across the length. It also takes longer — a few minutes for a long run — and you can leave the page while it works.
No. On Seedance 2.5 audio is generated with the picture at no additional charge, so it is on by default. That is specific to this model — on Kling 2.6 and Seedance 1.5 Pro, sound doubles the price.
480p and 720p. There is no 1080p or 4K on 2.5 — if you need those, Seedance 2.0 goes up to 4K but runs a maximum of 15 seconds. Length and resolution are the trade-off between the two generations.
A reference video of up to 30 seconds lends its motion and framing to the generation, and lowers the price by roughly 40% because the provider charges less when it has one. A reference audio track of up to 30 seconds drives speech, singing or rhythm. Both are optional and live under Advanced.
On this page, a specific one — 16:9 for landscape, 9:16 for vertical, 21:9 for cinematic. Adaptive is also offered, but it means "match the input material", and with a text-only prompt there is nothing to match, so it suits the image-driven pages better.
You are not charged. Credits are held when you submit and released back to your balance if the model fails or the provider never returns a result. Generated videos stay available for seven days, so download anything you want to keep.
Try Seedance 2.5 Text to Video
Thirty seconds, with sound, from a sentence. Draft it at 30 credits and see.