Seedance 2.5 API : Create AI Videos With Sound From Text, Images, Video or Audio
Try Seedance 2.5 online right here: write a prompt, add a first frame, reference images, videos or audio, or a clip to edit or extend, and get a video of 4 to 30 seconds with sound. Then call the same model from your product through the Seedance 2.5 API, with one key and one balance.
- Five modes
- Videos with sound
- Billed by output length
- Open Seedance 2.5 API
Features
Key Seedance 2.5 API Capabilities
Five ways to create with Seedance 2.5
The Seedance 2.5 API covers text to video, image to video, reference to video, video edit and video extension through one endpoint, POST /v1/tasks. Change the mode parameter and keep the rest of your integration as it is.
Video and sound in one pass
Every video from the Seedance 2.5 API comes with generated audio by default, so a clip is ready to post without a separate sound step. Set generate_audio to false when you only need the picture; the price per second stays the same.
4–30 seconds, 3 resolutions, 7 aspect ratios
Set duration anywhere from 4 to 30 seconds and render at 480p, 720p or 1080p in 21:9, 16:9, 4:3, 1:1, 3:4 or 9:16, or leave aspect_ratio on auto to follow your first frame or references.
First frame and last frame
In image to video, your first image opens the shot and an optional second image closes it, and Seedance 2.5 generates the motion in between. Use two landscape or two portrait images so the start and the end match.
References: images, videos and audio
Reference to video takes up to 30 images, 3 videos and 9 audio files in any mix, even audio alone. Put a character or product in a new scene, follow the moves and camera of a clip, or use your own track as the soundtrack.
Edit and extend existing footage
Video edit rewrites a clip of 4 to 30 seconds as your prompt says, for example swapping a product or changing the style, and keeps its length and aspect ratio. Video extension continues a clip forward or adds the moments before it.
Overview
What Is Seedance 2.5?
- Model ID
seedance-2.5- Modes
- 5
- Length
- 4–30s
- Resolution
- 480p / 720p / 1080p
- Aspect ratio
- auto · 21:9 · 16:9 · 4:3 · 1:1 · 3:4 · 9:16
- Images
- Up to 30
- References
- Up to 3 videos · 9 audio files
- Sound
- On by default, can be turned off
- Prompt
- Up to 8000 characters
- Price
- From $0.117 / s
Seedance 2.5 is an AI video model that generates the picture and the sound together. Give it a prompt, a first and last frame, reference images, clips or audio, or a video you already have, and it returns a video of 4 to 30 seconds at 480p, 720p or 1080p. Prompts can run up to 8,000 characters and name each file directly as @Image1, @Video1 or @Audio1, so one request can say exactly which character, move or soundtrack goes where.
On NanoAPI, Seedance 2.5 is one model with five modes. text_to_video generates from a prompt alone, image_to_video animates from a first frame (optionally with a last frame), reference_to_video builds a new scene from your reference images, videos and audio, video_edit rewrites a clip of 4 to 30 seconds, and video_extension continues a clip of up to 30 seconds forward or backward. Every mode uses the model value seedance-2.5; send resolution every time, aspect_ratio and duration in the modes that take them, and the video comes with sound unless you turn it off.
Use Seedance 2.5 online in the studio at the top of this page, then call the Seedance 2.5 API at https://api.nanoapi.ai/v1 from your own app. Both draw on the same prepaid balance at the same per-second prices, and a failed generation is never charged. Want to remake a viral video with one photo instead? Genjutsu 1.0, NanoAPI's motion transfer model, is fine-tuned from Seedance 2.5 for exactly that.
Modes
Seedance 2.5 Modes Compared
One endpoint, five modes: pick the one that matches what you start from. Every mode is billed per second of the video it generates, and video edit has a higher price per second.
| Mode | Input | Best for | 480p / s | 720p / s | 1080p / s |
|---|---|---|---|---|---|
Text to Videotext_to_video | A prompt (required) | Concept shots, B-roll and scenes you only have in words | $0.117 | $0.273 | $0.624 |
Image to Videoimage_to_video | A first frame and an optional last frame, both landscape or both portrait; optional prompt | Animating a photo or artwork, or moving between a fixed start and end | $0.117 | $0.273 | $0.624 |
Reference to Videoreference_to_video | Up to 30 reference images, 3 reference videos and 9 audio files, at least one; optional prompt | Keeping a character or product consistent, copying a move or camera, scoring a scene to your track | $0.117 | $0.273 | $0.624 |
Video Editvideo_edit | A clip of 4–30 s and a prompt; optional: up to 30 reference images | Swapping a product or object, or changing the style of footage you already have | $0.1404 | $0.2925 | $0.78 |
Video Extensionvideo_extension | A clip of up to 30 s, a direction (forward or backward) and a prompt | Continuing a shot, or adding what happens before it | $0.117 | $0.273 | $0.624 |
Live prices per second. Cost = output length × price; reference videos, reference audio and sound on or off cost the same, and video edit is billed by the actual output length. Failed runs are fully refunded.
How it works
How to Use the Seedance 2.5 API
- 1
Step 1: Get your API key
Sign in to NanoAPI with Google and create a key on the API Keys page of the console. The key authenticates every Seedance 2.5 API request, so keep it on your server.
- 2
Step 2: Try Seedance 2.5 in the studio
Test prompts, frames, references and edits in the studio above before you write code. It shows the cost before you generate: the length of the video × the price per second of the mode and resolution.
- 3
Step 3: Create a Seedance 2.5 video
Send POST /v1/tasks with model seedance-2.5, a mode and your prompt or media. Always send resolution, plus aspect_ratio and duration where the mode takes them (a video edit follows its clip); generate_audio is on unless you turn it off. The task is queued in under a second, and its cost is reserved within a few seconds.
- 4
Step 4: Get the result
Receive the finished task at your webhook_url or poll GET /v1/tasks/{task_id}. A Seedance 2.5 video usually takes a few minutes and comes back as a public video URL with sound, plus a silent, compressed preview.
Why NanoAPI
Why Use the Seedance 2.5 API on NanoAPI
One API for every mode
The Seedance 2.5 API uses the same POST /v1/tasks endpoint, task object and webhooks as every NanoAPI model, so a new mode or model never means a new integration.
Pay per second of output
The cost is the length of the video you get × the price per second of the mode and resolution. No subscription: top up from $10, and the balance never expires.
Sound and references cost nothing extra
Turning generate_audio on or off doesn't change the price, and neither do reference videos or reference audio. You pay for the seconds of video generated and nothing on top.
Failed runs are free
Each task reserves its cost when it's created and is charged only when the video is ready. If a generation fails, even because a file can't be downloaded, the full amount goes back to your balance. A video edit reserves 30 seconds first and returns what the finished video doesn't use.
Safe retries and spend limits
Send an Idempotency-Key and a resent request returns the task it already created instead of charging twice. Cap each key's hourly, daily or total spending in the console.
Try online, then integrate
The studio on this page runs the same Seedance 2.5 model at the same prices, so what you test is what your users get. The docs also come as one llms.txt file for AI coding assistants, and the public status page shows each model's success rate.
Use cases
What You Can Build With the Seedance 2.5 API
AI video generator apps
Turn prompts, photos and uploads into finished clips with sound in your own Seedance 2.5 video generator, creative tool or content platform built on the Seedance 2.5 API.
Ads and product videos
Generate launch videos, social ads and product showcases from a brief or a few product photos, then restyle an existing spot with video edit for every market.
Music and audio-led clips
Send a track as reference audio and use it as the soundtrack of a new video, for music promos, audio ads and branded sound.
Short dramas and stories
Keep a character consistent across shots with reference images, then carry each scene on with video extension or add what happened before it.
Social media content
Create vertical 9:16 or 3:4 videos for TikTok, Reels and Shorts, or square 1:1 posts, with the sound already in, straight from your scheduling or creator tools.
Automated video pipelines
Connect the Seedance 2.5 API to agents and content pipelines: create tasks in bulk, get the results by webhook and publish without manual steps.
FAQ
Seedance 2.5 API FAQ
Can’t find what you need? Write to us.
01What is the Seedance 2.5 API?
The Seedance 2.5 API lets your app generate videos with sound using the Seedance 2.5 model over HTTP. On NanoAPI you send POST /v1/tasks with model seedance-2.5 and one of five modes (text_to_video, image_to_video, reference_to_video, video_edit or video_extension), then get the finished video by webhook or by polling.
02How much does the Seedance 2.5 API cost?
You pay per second of the video generated: duration × the price per second of the mode and resolution. video_edit has a higher price per second and is billed by the length of the finished video: it reserves 30 seconds first and returns the rest. Reference videos, reference audio and generated sound don't change the price. The live prices are in the table above, on the pricing page and in GET /v1/products. Failed runs are not charged.
03Can I try Seedance 2.5 online before integrating?
Yes. Sign in and use the studio at the top of this page: it runs the same Seedance 2.5 model at the same prices as the API and draws on the same balance, so you can check results and costs before writing code. Top-ups start at $10.
04Does Seedance 2.5 generate audio?
Yes. Videos from the Seedance 2.5 API come with generated sound by default (generate_audio is true), and the price is the same with or without it. Set generate_audio to false for a silent video. The full video with sound is at outputs[].url, and preview_url is a compressed copy without sound.
05How long can Seedance 2.5 videos be, and which resolutions and aspect ratios are supported?
From 4 to 30 seconds, set in whole seconds with duration, at 480p, 720p or 1080p. Aspect ratios are 21:9, 16:9, 4:3, 1:1, 3:4 and 9:16, or auto, which follows the first frame or the references and falls back to 16:9. A video edit keeps the length and aspect ratio of the clip you edit, and a video extension keeps the aspect ratio of the clip it continues.
06What's the difference between video edit and video extension?
Video edit changes a clip you already have (4 to 30 seconds) as your prompt says, optionally with up to 30 reference images, and returns a new version of the same length. Video extension generates new footage that continues a clip forward or leads into it backward; the result is only the new part, so join it to the original yourself if you need one continuous video.
07How long does a Seedance 2.5 video take, and what if it fails?
Most videos take a few minutes. In production, pass a webhook_url to be notified once when the task finishes instead of polling. If a generation fails, the amount reserved for it goes back to your balance in full, and the task's error.type tells you why, for example a file that can't be downloaded or a reference video longer than 30 seconds.
08How is Seedance 2.5 different from Genjutsu 1.0?
Seedance 2.5 is a general video model with five modes, billed by the length of the video it generates. Genjutsu 1.0 is NanoAPI's own model, fine-tuned from Seedance 2.5 for motion transfer: one viral video and one photo give you the same moves, camera and timing with your person in them, billed by the source video's length with no extra fee for the reference video. Both use the same key and balance.
Start Creating With Seedance 2.5
Try Seedance 2.5 in the studio above, then use the Seedance 2.5 API to bring the same model into your product.