Animate one required start frame and an optional end frame. Seedance 2.5 uses Auto aspect ratio for this mode.
Upload Image
PNG, JPEG, or WebP, smaller than 30MB.
Optional ending image. The same file limits apply.
Audio will be generated with video
Neon Alley at Blue Hour
A cinematic tracking shot through a rain-soaked neon alley at blue hour, a red paper umbrella moving through reflected light, natural footsteps and distant city ambience, smooth camera motion, realistic physics.
Create longer AI videos from a prompt, guide motion with start and end frames, or combine image, video, and audio references in one omni workflow.
Move from a simple prompt to precise frame guidance or multi-modal references without leaving the generator.
Turn a written scene into a Seedance 2.5 clip with duration, resolution, aspect ratio, and sound controls.
Use this mode for concepts, cinematic shots, ads, and social clips.
Animate a required start image and optionally guide the transition toward a final frame.
This mode is useful when the opening composition or destination frame matters.
Create a new video guided by visual and audio references, or directly edit uploaded video footage.
Choose Reference Generation for a new result or Edit Video to modify footage. Edit Video uses Auto duration up to 30 seconds.
Generate a visual sequence with sound enabled, or switch sound off when you only need the video track.
The estimated credit total is shown before you submit.
Watch real Seedance 2.5 outputs and review the prompts and settings behind them.
A cinematic tracking shot through a rain-soaked neon alley at blue hour, a red paper umbrella moving through reflected light, natural footsteps and distant city ambience, smooth camera motion, realistic physics.
Up to 30 seconds
480p, 720p, and 1080p
Omni references
Audio references
Seedance 2.5 is an AI video model available on Flux Context for text-to-video, start-and-end-frame animation, omni-reference generation, and direct video editing.
Reference Generation and other generation modes support fixed output durations from 4 to 30 seconds. Edit Video uses Auto duration with a maximum of 30 seconds.
Seedance 2.5 supports 480p, 720p, and 1080p output in this generator.
You can add up to 30 images, 10 videos, and 10 audio files. Reference Generation accepts 2-30 second videos, while Edit Video requires 4-30 second videos. Audio files must be 2-30 seconds long, and video and audio each have their own 30-second total limit.
Yes. Open Omni Reference, choose Edit Video, upload at least one video, describe the changes, and generate. The aspect ratio and output duration are set to Auto, with a maximum output of 30 seconds.
Edit Video credits are estimated from the measured length of your uploaded videos and the expected output length. After completion, the final amount is settled from the actual input and output video duration, with any difference returned or additionally charged. Failed tasks are refunded through the normal task flow.
Yes. Omni-reference mode accepts audio-only reference input, so an image or video reference is not required when audio is supplied.
Active tasks can be resumed, and completed outputs remain available from My Creations when the task is saved to your account.
Choose a generation or editing workflow, review the duration and credit estimate, and submit your video task.
