Seedance 2.5
Seedance 2.5 creates short videos from a prompt, first and last frames, or large image, video, and audio reference sets. Use 480p or 720p output, native audio, official aspect ratios, and up to 30 seconds for richer creative direction.
Curated examples
Output Details
Model type
Prompt, frames, and references to video
Resolution
480p / 720p
Duration
4-30 seconds
Aspect ratio
16:9 / 4:3 / 1:1 / 3:4 / 9:16 / 21:9 / Adaptive
Reference inputs
Images (30) / Videos (10) / Audio (10)
Audio
Optional generated audio
Estimated time
About 90-240 seconds
Credits
Based on resolution, duration, and reference video input
Key Features
Three input modes
Generate from a prompt, guide the motion with first and last frames, or combine reference images, videos, and audio.
30-second clips
Create longer short-form videos from 4 to 30 seconds for narrative beats, product demos, and social sequences.
Native audio control
Enable or disable generated audio while keeping the video input workflow in one place.
Large reference sets
Use up to 30 images, 10 videos, and 10 audio references when the scene needs richer visual or sound direction.
Best Uses
Product storytelling
Animate product details, transitions, and camera movement across a longer commercial-style clip.
Reference-led scenes
Combine visual and audio references to keep characters, materials, locations, and rhythm consistent.
Social sequences
Create vertical, square, landscape, or adaptive clips for short-form campaigns and concept tests.
FAQ
Can Seedance 2.5 generate without a prompt?
Yes. Frames or reference media can carry the visual direction, so the prompt can be empty when the uploaded inputs are sufficient.
Can I mix frame inputs and reference media?
No. Choose Frames or References for each generation. The two input modes are mutually exclusive.
How long can a Seedance 2.5 clip be?
PixPilot exposes 4 to 30 seconds. Intelligent duration is not offered because the final cost cannot be known before the provider finishes.
How are Seedance 2.5 credits calculated?
Credits use the selected resolution, duration, and whether a reference video is attached. The estimate is rounded up to a multiple of five.
Related Models
Seedance 2.0
VideoFlexible video generation with prompts, frames, references, and optional audio.
Hailuo 2.3
VideoRealistic text-to-video and image-to-video generation at 768p or 1080p.
Hailuo 02
VideoPhysics-focused video generation with prompt, image, and first-to-last-frame modes.
Kling v3 omni
VideoPrompt, image, frames, and reference-guided video generation with optional audio.
Veo 3.1
VideoHigh-fidelity AI video generation with native audio, image animation, frame control, and reference images.
Ready to try Seedance 2.5?
Open the workspace and start creating with Seedance 2.5.