Generate clips with fluid motion and audios for creatives
Seedance 2.5 Flash Text To Video writes a 4 to 30 second clip from a prompt.
The output is 480p or 720p. Five seconds and 720p are the starting point. Native sound is recorded into the same file unless you turn it off. The frame can be 16:9, 9:16, 4:3, 3:4, 1:1, or 21:9, and 16:9 is the default. This page does not render 1080p or 4K. Those resolutions belong to the flagship Seedance 2.5 pages.
The request on this page is the prompt. A still that has to be the first frame belongs on the image page. Reference images, clips, and audio belong on the reference page.
Use the lighter 480p pass while you search for the shot, then 720p for the keep. For developers, the same controls are on the RunComfy model page and over the HTTP API, so you do not host or scale the model yourself.
Ideal for: Social Drafts | Prompt-Only Scenes | Fast Motion Tests
The prompt is the only required field. Aspect ratio, duration, resolution, and sound are optional.
| Parameter | Required | Type | Default | Range / Options | Description |
|---|---|---|---|---|---|
prompt* | Yes (*) | string | — | Up to 20,480 characters | Scene, motion, camera, and sound. |
aspect_ratio | No | string | 16:9 | 16:9, 9:16, 4:3, 3:4, 1:1, 21:9 | Output frame. |
duration | No | integer | 5 | 4-30 seconds | Generated length. |
resolution | No | string | 720p | 480p, 720p | 480p is the draft pass. 720p is the keep. |
generate_audio | No | boolean | true | true / false | Native sound in the same file. |
A last frame is not a control on this page. If the clip must open on a still, use the image page. If the shot needs reference images, clips, or audio, use the reference page.
Name one camera move, then the action in time order, then two or three sounds. A second camera move in the same sentence tends to split the shot. Turn sound off when the file will be scored later. The rate stays the same whether the soundtrack is on or off.
Write the subject first, then one action, then the camera. Five seconds is enough to see whether the move holds. Ten seconds is the usual length for a social cut. Thirty seconds is the longest file this page will make, and it costs the same per second as a short one. Keep the ask to a single location. A jump to a second place in the same prompt tends to smear the middle of the clip. If the first pass is close, change one thing at a time: the verb, the camera, or the sound.
Seedance 2.5 Flash Text To Video bills the generated length. The rate is $0.092 per second at 480p and $0.204 per second at 720p. Sound does not change the rate. The hold at submit uses the duration you set. The final charge uses the length of the file that comes back.
| Resolution | Per second |
|---|---|
| 480p | $0.092 |
| 720p | $0.204 |
| Example | Counted seconds | Charge |
|---|---|---|
| 5 s at 720p | 5 | $1.020 |
| 5 s at 480p | 5 | $0.460 |
| 4 s at 480p | 4 | $0.368 |
| 10 s at 720p | 10 | $2.040 |
| 30 s at 720p | 30 | $6.120 |
A 5-second 720p file is the usual first charge. Drop to 480p, or shorten the clip, while you are still changing the prompt.
Generate clips with fluid motion and audios for creatives
Turn reference images and a prompt into short video with synced audio.
Transform images into smooth 1080p motion clips with synced audio and rapid rendering.
Create cinematic clips in seconds with Veo 3.1 Fast, built for instant text-driven motion and creative control.
Guide cinematic clips with up to seven images and precise prompts.
FLUX 3 Max Text To Video turns a prompt into a short clip.
Seedance 2.5 Flash Text To Video writes a 480p or 720p clip from a prompt. It fits social drafts and prompt-only scenes.
This page outputs 480p and 720p only, for 4 to 30 seconds, with native audio. It does not render 1080p or 4K. The request is a prompt. Reference images, clips, and audio belong on the reference page.
The required field is the prompt. Optional fields are aspect ratio, duration from 4 to 30 seconds, resolution, and generate audio. The default is 5 seconds at 720p in 16:9 with sound on.
No. The rate stays the same whether native sound is on or off. The charge follows the generated length.
The rate is $0.092 per second at 480p and $0.204 per second at 720p, billed on the generated length. A 5-second 720p clip is $1.020. A 4-second 480p clip is $0.368.
Yes. Shape the prompt and settings on the RunComfy model page, then send the same parameters to the RunComfy API. Both paths consume usd. The page does not promise a response-time SLA.
Use the image page when the clip must open on a specific still. Use the reference page when the shot needs reference images, clips, or audio. Use this page when the shot starts from the prompt.
RunComfy is the premier ComfyUI platform, offering ComfyUI online environment and services, along with ComfyUI workflows featuring stunning visuals. RunComfy also provides AI Models, enabling artists to harness the latest AI tools to create incredible art.





