GPT Image 2.5 Flare & Sunburst are live on EvoLinkTry GPT Image 2.5

Seedance 2.5 Flash API

Use five Seedance 2.5 Flash routes through EvoLink’s unified video API: text-to-video, first/last-frame image-to-video, multimodal reference-to-video, video edit, and video extend. Check model IDs, limits, and billed seconds before integrating.

ByteDance / BytePlusVideo GenerationAvailable
From ~$0.056 / s for 480p output with a reference video
Text-to-VideoImage-to-VideoReference-to-VideoVideo EditVideo ExtendVideo + Audio
Production routeLive
Model Highlights
4-30s clips · 480p / 720p · image, video, and audio references
Use Cases
Ad variants, social clips, quick edits, and video extensions
Input
Text, image, video, or audio
Output
Video with synchronized audio

Try Seedance 2.5 Flash before you integrate the API.

Test Seedance 2.5 Flash output, estimate credit usage, and generate a sample before you formally integrate.

Create a Sample

Pick a video mode, enter your prompt, and set the parameters.
Mature ModeAPI only

Off here. The API still supports it — send content_filter: false to relax content restrictions (billed at +10%). Illegal and prohibited content is always blocked.

Duration5s
Quality
Aspect Ratio
Generate Audio
Synchronized audio, no extra charge.
Web Search
Optional, billed when used.
Output Format
Estimated: ~$0.97166 cr
720p × 5s
Ready

Pick a mode, describe the shot, and hit Generate — your video will appear here.

History

Results are kept for 24 hours — download anything you want to keep.

Your generation history will appear here

Know your Seedance 2.5 Flash testing cost before you add credits.

Start from the cost of a single sample and pick a testing budget. Expand for detailed rates.

First Test Cost

Text-to-Video, 4s at 480p — the cheapest possible request, no reference files
Minimum charge
Credits24.552

One 4s 480p sample

Approx. Cost$0.362

About 26 tests with $10 in credits.

In Playground, choose Text to Video, drag Duration to 4s, and pick 480p. Use 480p for lower-cost prompt iteration and 720p for final output.

Testing Budget Guide

At the first-test settings: 4s at 480p
Add Credits
$10
About 26 tests

Good for a first validation.

$50
About 138 tests

Good for prompt iteration.

$100
About 276 tests

Good before production integration.

Seedance 2.5 Flash Cost Examples

Quality checkText-to-Video · 720p · 5s
~$0.971
~66 cr
Reference with video720p · 5s output + 3s reference · 10s billed
~$1.183
~80.388 cr
Video edit720p · 10s source · 20s billed
~$2.365
~160.776 cr

Estimated cost per run for the listed parameters. Final cost follows the mode, resolution, and reference video length.

Seedance 2.5 Flash Model Pricing

Text-to-Video
No video input · Output video seconds
480p $0.091/s · 6.138 cr/s
720p $0.195/s · 13.2 cr/s
Image-to-Video
No video input · Output video seconds
480p $0.091/s · 6.138 cr/s
720p $0.195/s · 13.2 cr/s
Reference-to-Video
Image or audio reference · Output video seconds
480p $0.091/s · 6.138 cr/s
720p $0.195/s · 13.2 cr/s
Reference-to-Video
With reference video · Output + max(reference, output) seconds
480p $0.056/s · 3.7488 cr/s
720p $0.119/s · 8.0388 cr/s
Video Edit
Input video required · Output + max(reference, output) seconds
480p $0.056/s · 3.7488 cr/s
720p $0.119/s · 8.0388 cr/s
Video Extend
Input video required · Output + max(reference, output) seconds
480p $0.056/s · 3.7488 cr/s
720p $0.119/s · 8.0388 cr/s

Billed per second; audio adds no extra cost. Without video input, billed seconds = output duration. With video input (Reference-to-Video with a video, Video Edit, Video Extend), the lower video-input rate applies and billed seconds = output + the longer of the total input video length or the output, rounded up to 0.1s. Auto duration (duration: -1) and Video Edit settle on the actual output length. Requests with content_filter set to false are billed at +10%. Failed tasks are not charged.

Auto duration: the model picks the length, so credits are reserved at the 30s cap — about 396 credits (~$5.824) at 720p. You are only charged for the seconds actually generated; the rest is refunded when the task finishes.

ItemUse CasePriceCredits
Web SearchPer searchUse when the prompt needs current events, products, or real-world references.$0.0006 / search0.04 cr

Most first tests only pay the video generation fee. Logged-in users can view live pricing in the console.

What is the Seedance 2.5 Flash API?

Access the Seedance 2.5 Flash API through EvoLink for text-to-video, first/last-frame image-to-video, reference-to-video, video editing and video extension. Generate 4–30 second clips at 480p or 720p with synchronized audio, or let the model pick the length and aspect ratio automatically; edits follow the input video. Requests run asynchronously: submit a task, keep its ID and retrieve the final result. Choose the route by your inputs and compare the cost of accepted shots for ad variants, social clips and quick edits.

Input
Text · Image · Video · Audio
Output
Video with synchronized audio (mp4 or mov)
Duration
4–30 seconds or auto (edit follows the input)
Resolution
480p · 720p
References
Up to 10 images, 5 videos, 5 audio tracks

Seedance 2.5 Flash API code example and error handling

This example shows the shortest runnable flow: create a task, save the returned task ID, then poll or wait for an HTTPS callback. Open the API tab for the complete parameter and response reference.

View complete API docs
cURL
curl -X POST https://api.evolink.ai/v1/videos/generations \
  -H "Authorization: Bearer $EVOLINK_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "seedance-2.5-flash-text-to-video",
  "prompt": "A paper boat drifts down a rain-soaked city street at dusk, neon reflections ripple across the water, slow tracking shot",
  "duration": 5,
  "quality": "720p",
  "aspect_ratio": "adaptive",
  "generate_audio": true,
  "content_filter": true
}'

# Save the returned task ID, then query:
curl https://api.evolink.ai/v1/tasks/{task_id} \
  -H "Authorization: Bearer $EVOLINK_API_KEY"

Seedance 2.5 Flash parameter validation failed

Check that duration is an integer from 4 through 30 or -1, quality is 480p or 720p, and the asset counts match the route (up to 10 images, 5 videos, 5 audio tracks). 1080p and Draft mode are not available.

Seedance 2.5 Flash authentication or balance issue

Check the Authorization bearer token and confirm the available balance and task billing in the console.

Seedance 2.5 Flash content or asset rejected

Review likeness rights, brand assets, sensitive content, and whether each asset is a reachable HTTP(S) URL in a supported format.

Seedance 2.5 Flash task failed or timed out

A task that reaches a final failed state is not charged. Keep the task ID, inspect the final state, and identify whether the cause is validation, review, or execution before retrying.

Seedance 2.5 Flash callback not received

callback_url must be a public HTTPS URL, not localhost or a private network address. Always keep the task ID as a polling fallback.

Two ways to use Seedance 2.5 Flash: EvoLink API or Agent

Use the EvoLink API for product integration and batch jobs, or call it from Codex, Claude, or Gemini for fast creative and development workflows. Both paths share the same EvoLink API key, balance, model routes, and task history.

Option 1

Integrate Seedance 2.5 Flash with the EvoLink API

Best for: product backends, batch jobs, automated workflows

Call EvoLink’s unified video API from your server and control the model ID, parameters, task queue, callbacks, and result storage.

  1. 1Validate output and cost with a real brief in Playground
  2. 2Create an EvoLink API key in the console
  3. 3Choose the model ID that matches your input
  4. 4Submit the task and retrieve the result by polling or HTTPS callback
Option 2

Call Seedance 2.5 Flash with an Agent

Best for: creative and development tasks in Codex, Claude, and Gemini

Give the Agent your output goal, assets, and acceptance criteria. It can choose the route, assemble the request, track the task, and return the result without requiring you to hand-code every step.

  1. 1Set EVOLINK_API_KEY in your local environment; never put it in code or a prompt
  2. 2Describe the output goal, references, duration, and resolution
  3. 3Ask the Agent to call a Seedance 2.5 Flash route and save the task ID
  4. 4Let the Agent poll the final state, download the result, and report failure reasons

Seedance 2.5 Flash API features and available generation modes

Pick the route by what you already have: a script, a first frame, a set of references, or a video to change or continue. Every route posts to the same endpoint; the model field selects the route.

Seedance 2.5 Flash Text-to-Video API

seedance-2.5-flash-text-to-video

Start with a script or prompt when no source footage is needed. Generate 4–30 seconds (or auto length) with synchronized audio in six aspect ratios or adaptive; choose 480p or 720p and bill output duration at the no-video-input rate.

Seedance 2.5 Flash Image-to-Video API

seedance-2.5-flash-image-to-video

Animate one image as the first frame, or use two images to constrain the first and last frames. The output follows the first frame’s aspect ratio. Generation lasts 4–30 seconds and uses the no-video-input rate.

Seedance 2.5 Flash Reference-to-Video API

seedance-2.5-flash-reference-to-video

Combine up to 10 images, 5 videos and 5 audio tracks; audio can be used on its own. Assign each asset a role using @image1, @video1 and @audio1. Video references are 2–30 seconds each, up to 30 seconds combined, and change both the rate and billed duration.

Seedance 2.5 Flash Video Edit API

seedance-2.5-flash-video-edit

Edit a video with a prompt — replace a subject, add or remove objects, or restyle a scene — with optional image and audio references. The video to edit is 4–30 seconds; output length and aspect ratio follow it. Billed seconds are the output plus the input video length at the video-input rate.

Seedance 2.5 Flash Video Extend API

seedance-2.5-flash-video-extend

Continue a video forward or backward with a prompt, plus optional image and audio references. Set the new segment to 4–30 seconds or auto; the aspect ratio follows the input video. Billed seconds are the output plus the longer of the input video length or the output, at the video-input rate.

4–30 second Seedance 2.5 Flash shots

Keep the hook, main action, and ending in one clip at 480p or 720p.

Multimodal references in Seedance 2.5 Flash

Up to 10 images, 5 videos, and 5 audio tracks, referenced in the prompt as @image1, @video1, @audio1.

Synchronized audio in Seedance 2.5 Flash output

Dialogue, sound effects, and music are generated with the video at no extra charge.

Asynchronous Seedance 2.5 Flash tasks

Manage jobs with a task ID, polling, and HTTPS callbacks.

Seedance 2.5 Flash API key request parameters

modelstring

Selects the route — one of the five model IDs above.

promptstring

Required on every route, up to 10,000 tokens. Refer to assets as @image1, @video1, @audio1.

durationinteger

Integer from 4 through 30, default 5, or -1 to let the model pick the length (billed by actual output). Video Edit always uses -1: the output follows the input video.

qualitystring

480p or 720p, default 720p. Billing follows the quality tier.

aspect_ratiostring

adaptive (default), 16:9, 9:16, 1:1, 4:3, 3:4, or 21:9 on Text-to-Video and Reference-to-Video. Other routes follow the first frame or the input video.

image_urlsstring[]

Image-to-Video: 1–2 first/last-frame images. Reference-to-Video, Video Edit and Video Extend: up to 10 images.

video_urlsstring[]

Reference-to-Video: up to 5 videos. Video Edit and Video Extend: 1–5 videos, the video to edit 4–30s. Each video 2–30s, 30s combined.

audio_urlsstring[]

Up to 5 audio tracks, 30s combined, for voice, music, or sound reference.

generate_audioboolean

Default true, no extra charge. On Video Edit and Video Extend, false means no new audio is generated.

content_filterboolean

Default true. false relaxes content restrictions and is billed at +10%; it cannot be combined with web search.

output_formatstring

mp4 (default) or mov for higher color fidelity in editing workflows, at no extra charge.

model_params.web_searchboolean

Text-to-Video only. Searches the web when the prompt needs current information; billed per search.

callback_urlstring

Public HTTPS URL that receives the final task state; keep the task ID as a polling fallback.

Unsupported values return 400 with the field name. Media fields take HTTP(S) URLs or asset:// IDs. Full constraints and response fields are in the API Reference.

What can you build with Seedance 2.5 Flash?

Choose a workflow and test it with your own brief in Playground. Check the subject, motion, audio and required output before accepting a shot, and compare retries and total cost. The workflows below are evaluation plans, not recorded Flash outputs.

Ad variants with Seedance 2.5 Flash

Generate several hooks, angles, and aspect ratios from one product brief, then keep the variants that test best.

Social clips from a first frame

Animate a product shot or key visual with Image-to-Video, or lock both ends of the shot with first and last frames.

Character and product consistency

Feed reference images, a motion clip, and a voice or music track to Reference-to-Video so the subject stays on-model.

Quick edits to existing footage

Swap a wardrobe item, remove an object, or restyle a scene with Video Edit while the source length and framing stay the same.

Seedance 2.5 Flash API limits and model selection

Flash is not a drop-in replacement for every Seedance task. Use the same brief, input assets, and acceptance criteria to compare usable-shot rate, reruns, and total project cost.

Decision factorSeedance 2.5 FlashSeedance 2.5Seedance 2.0 Mini
Best suited toFrequent 4–30s drafts, ad variants, quick edits, and extensions at 480p or 720pProduction shots up to 1080p, larger reference sets, and draft-to-final workflowsLow-cost short clips and prompt iteration on the Seedance 2.0 generation
Duration4–30 seconds, or auto4–30 seconds, or auto4–15 seconds
Resolution480p / 720p480p / 720p / 1080p480p / 720p
RoutesText, image, and reference to video, video edit, and video extendThe same routes plus draft to finalText, image, and reference to video
Cost decisionMeasure cost per usable shot when you iterate a lot or edit existing footageMove a shot to 2.5 when it needs 1080p, more references, or a draft-to-final workflowKeep it for short clips if your current 2.0 baseline already meets the brief

Key details before you integrate the Seedance 2.5 Flash API

Three facts that shape the integration: routing, task lifecycle, and billing.

One endpoint for five Seedance 2.5 Flash routes

Every route posts to /v1/videos/generations. The model field selects the route; unsupported values return 400 with the field name, so check the API Reference before reusing a request.

Seedance 2.5 Flash tasks are asynchronous

A request returns a task ID immediately. Poll every 5–10 seconds or register a public HTTPS callback_url, and keep the task ID either way. Result links are valid for 24 hours. A task whose final state is failed is not charged.

Seedance 2.5 Flash billing follows billed seconds

Without video input, you pay for the output seconds at the 480p or 720p rate. With video input (references, edit, extend), billed seconds are the output plus the longer of the input video length or the output, at the lower video-input rate. Auto duration reserves credits at the 30-second cap and settles on the actual output.

Seedance 2.5 Flash billed-seconds examples

These examples explain billing quantity, not a fixed quote. Apply the selected 480p or 720p rate in the existing Pricing section; the starting video-input rate does not apply to every output second.

5-second text-to-video

5 output seconds = 5 billed seconds at the no-video-input rate. The same basis applies to image-to-video and references without video.

5-second output with a 2-second reference

5 + max(2, 5) = 10 billed seconds at the video-input rate. Billed seconds are rounded up to 0.1 seconds; a 12.1-second reference would make this 5 + 12.1 = 17.1 billed seconds.

Editing a 5-second video

About 5 output seconds + 5 input seconds = 10 billed seconds at the video-input rate. The output follows the input video, so the final amount settles on the actual output length.

Why access Seedance 2.5 Flash through EvoLink?

The hard part of a production Seedance integration is not one request. It is managing model choice, cost, references, asynchronous jobs, callbacks, and failures after launch. EvoLink brings these into one video API and console so teams can test first, integrate second, and route by usable-shot rate and total project cost.

Create Seedance 2.5 Flash API Key

Manage Seedance 2.5 Flash and other video models with one API key

All five Seedance 2.5 Flash routes share EvoLink authentication, balance, and task infrastructure. Your server stores one EVOLINK_API_KEY and selects the model ID for each input.

Keep your EvoLink API key and asynchronous task workflow when evaluating another video model. Select its model ID and check that route’s input fields, duration, resolution, and billing before sending a request.

Test in Playground before integrating Seedance 2.5 Flash into production

Use a real ad brief, product image, reference clip, and audio asset in Playground before writing your production integration. That is more reliable than judging the model with one demo prompt.

Track usable-shot rate, reruns, and manual rework—not only price per second—before you lock the route, prompt structure, and acceptance criteria.

Failed Seedance 2.5 Flash tasks are not charged, and billing stays auditable

EvoLink does not charge a task whose final state is failed. Save the task ID and verify the final status, failure reason, and billing record in the console.

Production systems should use polling or a public HTTPS callback, handle missing or duplicate callbacks and timeouts, and retry only after identifying the cause.

Seedance 2.5 Flash API and Agent workflows share one foundation

Product backends can run batch tasks through the EvoLink API, while creative and development teams can ask Codex, Claude, or Gemini to assemble requests, track status, and return results.

Both methods share the API key, routes, balance, and task history. Copy the Agent prompt on this page, fill in the generation goal, and paste it into your Agent.

Seedance Model Family

Compare Seedance routes under one EvoLink account and API key, then choose by shot length, resolution, reference workflow, and cost.

Compare all Seedance models
Seedance 2.5 API

Seedance 2.5

The full Seedance 2.5 route with 1080p output, larger reference sets, and a draft-to-final workflow for production shots.

View API
Seedance 2.0 API

Seedance 2.0

A mature Seedance production baseline for shorter shots, multimodal references, and existing 2.0 workflows.

View API
Seedance 2.0 Mini API

Seedance 2.0 Mini

A lightweight Seedance 2.0 route for low-cost drafts, prompt iteration, and frequent batch jobs.

View API

Other video models on EvoLink besides Seedance 2.5 Flash

When Seedance 2.5 Flash is not the right default, use the same brief to compare resolution, shot length, audio, motion stability, and total project cost.

Kling 4.0 Flash API

Kling 4.0 Flash

Kling’s lower-cost 720p tier with text, first-frame, and reference-to-video plus video edit.

View API
MiniMax H3 API

MiniMax H3

For detailed 2K clips of 4–15 seconds; compare its usable-shot rate with Seedance 2.5 Flash on the same brief.

View API
Veo 3.1 API

Veo 3.1

Google’s video generation model for cinematic visuals and native-audio workflows.

View API
Tencent Video Upscale API

Tencent Video Upscale

When a Flash shot already has usable motion and composition but looks soft, enhance clips up to 60 seconds to 1080p, 2K, or 4K.

View API

Seedance 2.5 Flash API FAQ

Questions about model IDs, pricing, limits, editing, and how Flash differs from Seedance 2.5.

Is the Seedance 2.5 Flash API available on EvoLink?

Yes. Seedance 2.5 Flash is available through five routes: seedance-2.5-flash-text-to-video, seedance-2.5-flash-image-to-video, seedance-2.5-flash-reference-to-video, seedance-2.5-flash-video-edit, and seedance-2.5-flash-video-extend. You can test it in Playground first and then call it with the same EvoLink API key.

How is Seedance 2.5 Flash different from Seedance 2.5?

Flash supports 480p and 720p, up to 10 reference images, 5 videos, and 5 audio tracks, and no Draft mode. Request fields, auto duration, adaptive aspect ratio, video edit, and video extend work the same way as on Seedance 2.5. Seedance 2.5 adds 1080p, larger reference sets, and a draft-to-final workflow.

How is Seedance 2.5 Flash billed?

Per billed second. Without video input you pay for the output seconds; with video input, billed seconds are the output plus the longer of the input video length or the output, at a lower video-input rate. Auto duration and Video Edit settle on the actual output length. See the Pricing section for current rates.

Are failed Seedance 2.5 Flash tasks charged?

No. A task whose final state is failed is not charged. Keep the task ID and check the final state, failure reason, and billing record in the console.

What durations, resolutions, and aspect ratios does Seedance 2.5 Flash support?

Duration is an integer from 4 through 30 seconds (default 5), or -1 to let the model choose. Quality is 480p or 720p (default 720p). Text-to-Video and Reference-to-Video accept adaptive (default), 16:9, 9:16, 1:1, 4:3, 3:4, and 21:9; Image-to-Video follows the first frame, and Video Edit and Video Extend follow the input video.

How many references can I use with Seedance 2.5 Flash?

Reference-to-Video accepts up to 10 images, 5 videos (2–30 seconds each, 30 seconds combined), and 5 audio tracks (30 seconds combined); audio can be used on its own. Video Edit and Video Extend take 1–5 videos plus up to 10 images and 5 audio tracks. Media can be HTTP(S) URLs or asset:// IDs.

What happens to audio when I edit or extend a video with Seedance 2.5 Flash?

generate_audio defaults to true, so the result comes with newly generated audio. When you set it to false, no new audio is generated. Test with your own footage if the original soundtrack matters to your workflow.

Can I use Seedance 2.5 Flash through an Agent?

Yes. Set EVOLINK_API_KEY in Codex, Claude, or Gemini, copy the Agent prompt on this page, and let the Agent choose the Seedance 2.5 Flash route, submit the task, poll the status, and return the video. Keep the key in an environment variable, not in the prompt.

Can the Seedance 2.5 Flash API generate native 1080p or 4K video or use Draft mode?

These EvoLink Flash routes support 480p and 720p. They do not support native 1080p, native 4K, or Draft mode. Evaluate the Seedance 2.5 page for 1080p and the draft-to-final workflow; upscaling a Flash clip is a separate processing step.

Can I migrate to Seedance 2.5 Flash by changing only the model ID?

Reuse your EvoLink API key, asynchronous task flow, and request fields, then change the model ID. Flash uses quality 480p or 720p and accepts up to 10 images, 5 videos, and 5 audio tracks; draft is not available. Check the API Reference before reusing a Seedance 2.5 request with 1080p or larger reference sets.