Short answer: Our Seedance 2.0 API lets you call Seedance 2.0 Pro, Fast and Mini, plus Seedance 2.5, with one API key. To choose a model and integrate it, retrieve its supported options with GET /models, preview the cost with POST /quotes, and submit an asynchronous generation request.
Which Seedance models can I call through your API?
Our API provides a shared request structure for choosing models, supplying prompts and image inputs, and controlling supported generation parameters. You can start with a Seedance 2.0 model and evaluate a 2.5 upgrade through the same integration. The important distinction is between a common request structure and model-specific capabilities: each model accepts only the options it advertises.
Before carrying an existing integration over to another model, explore Seedance 2.5 AI video generator and define the scene and output requirements you will check against the API.
The following comparison shows the public Seedance model IDs and where they fit into model selection. For Seedance 2.0 models, retrieve resolutions, durations and other supported settings through GET /models rather than assuming that options transfer between variants.
| Model ID | What to use it for | How to choose supported settings |
|---|---|---|
seedance-2.5 | The flagship choice: videos up to 30 seconds at up to 1080p, and the natural upgrade from a Seedance 2.0 workflow. | Check its advertised options before changing an existing request. |
seedance-2.5-lite | Planning a workflow around the documented output-second pricing structure. | Check supported inputs and include reference-image charges in your quote. |
seedance-2.0-pro | Integrating the Pro variant of Seedance 2.0. | Retrieve the model’s supported modes and parameters. |
seedance-2.0-fast | Integrating the Fast variant of Seedance 2.0. | Use its own capability response when constructing requests. |
seedance-2.0-mini | Integrating the Mini variant of Seedance 2.0. | Check its advertised settings before submitting a task. |
seedance-1.5-pro | Selecting Seedance 1.5 Pro within the shared API integration. | Retrieve its capabilities separately from the newer models. |
The same key also covers video models kling-3.0 and veo-3.1-fast, and image models gpt-image-2, nano-banana-2 and nano-banana-pro. For the complete model list and pricing rules, use the API models and pricing reference. If Mini is your intended starting point, see the Seedance 2.0 Mini page alongside its API capability response.
Comparing older models? Read Seedance 1.5 Pro vs 2.5 for model IDs and free credits on sign-up.
How does Seedance 2.0 API pricing work?
Seedance 2.0 API pricing depends on the selected model and request parameters, with prices following the website. The practical way to budget a request is to quote its exact configuration. POST /quotes returns cost_credits without generating an output or charging for generation.
Seedance 2.5 Lite has a documented rate of 10 credits per output second. Its first five reference images are included; each additional image costs 7 credits. That gives you a clear basis for estimating Lite requests, while a quote remains useful for checking the complete configuration before submission.
Build cost handling into the integration in this order:
- Read the website credit balance with
GET /credits/balance. - Prepare a request using supported model options.
- Send the configuration to
POST /quotesand inspectcost_credits. - Set
max_cost_creditson the generation request to enforce your acceptable ceiling.
If the quote exceeds max_cost_credits, the request is rejected before generation. This makes the field useful when your application lets users change duration, resolution or image inputs after an initial estimate. Quote the final configuration and carry an explicit budget into submission.
Credits are reserved before submission, and a verified failure refunds once. Queries, quotes and downloads cost no generation credits. Keep credit reservations and refunds in mind when displaying available spending capacity, and use the balance endpoint when your application needs to read the account balance.
What limits should I plan for before integrating?
Account limits are shared across API keys. Creating additional keys does not provide separate generation capacity or separate daily allowances, so your application should coordinate work at the account level.
- Active generation: 1 active API generation.
- Generation and upload requests: 10 per minute and 1,000 per day.
- Reads and quotes: 60 per minute and 20,000 per day.
- API generation credit budget: 2,000 credits per day.
- Image uploads: 100 files and 500 MiB per day.
Daily windows reset in UTC. Use that reset boundary when displaying daily usage or planning queued work. A local calendar day may differ from the account’s API usage window.
Because generation is asynchronous, coordinate submissions with task status. Polling also needs to fit within the read allowance, while image-based workflows need to account for both upload requests and daily file limits. For additional planning context, read the guide to Seedance 2.5 API pricing and limits.
How do I get a Seedance API key and make my first call?
Start by signing up, then follow the API integration documentation to create and configure your Seedance API key. New accounts get free credits on sign-up. Use the documented authentication instructions for your requests, and keep the key in your application’s server-side configuration.
Most users integrate on their own by following the /docs/api documentation; if you need more permissions or have special requirements, contact support.
For a first integration, work through the following sequence:
- Discover capabilities. Call
GET /modelsand select a public model ID. Read its supported modes, resolutions, durations, aspect ratios and image limits. - Prepare inputs. Write the prompt and, when using images, supply uploaded image URLs through
inputs. - Choose supported parameters. Include only settings advertised for the selected model and mode.
- Check the cost. Read the balance, request a quote and choose a suitable
max_cost_credits. - Submit and retrieve. Follow the documented generation submission flow, poll task status and save signed outputs before expiry.
This short JSON template illustrates a Seedance 2.0 Fast API request. Replace the mode placeholder with a supported value returned by GET /models, and add any parameters required for that configuration before quoting or submitting it.
{
"model": "seedance-2.0-fast",
"mode": "REPLACE_WITH_SUPPORTED_MODE",
"prompt": "A slow camera movement through a sunlit garden."
}The available parameter fields include parameters.resolution, parameters.aspect_ratio and parameters.duration_seconds. For image workflows, parameters.image_mode can specify first-frame, first-last-frame or reference-image where supported. Include parameters.generate_audio only when the model advertises audio control.
Unsupported models and options are rejected, and unknown fields are rejected too. Construct requests from the capability response instead of passing every possible parameter to every model. This is especially useful when a single application allows users to switch models.
Why should I upgrade from Seedance 2.0 to Seedance 2.5?
Seedance 2.5 supports videos up to 30 seconds and 1080p, making it an upgrade to evaluate when those output requirements matter to your application. You can use the same API key and request shape, changing the model ID to seedance-2.5.
Keep capability discovery in the migration process. Retrieve the new model’s supported settings, adjust the request parameters and obtain a fresh quote before generation. The shared request structure simplifies the integration work, but you should still validate the configuration against the selected model.
A useful evaluation starts with prompts and inputs from your existing workflow. Review the resulting outputs against your application’s requirements, and consult the video showcases for examples. Retain the same cost controls, account-level scheduling and output-saving steps when moving to 2.5.
Frequently asked questions
Can I use one Seedance API key for 2.0 and 2.5?
Yes, one key covers Seedance 2.0 Pro, Fast and Mini, plus Seedance 2.5 and 2.5 Lite. Select the model through the model field and retrieve its supported settings before submission. Account limits remain shared across keys.
How much does the Seedance 2.0 API cost?
The cost depends on the model and request parameters. Use POST /quotes to retrieve cost_credits without generating or charging, then set max_cost_credits to prevent generation above your chosen ceiling.
What model ID do I use for the Seedance 2.0 Fast API?
Use seedance-2.0-fast. Get its supported modes, resolutions, durations, aspect ratios and image limits from GET /models. Build the request from that response rather than reusing another model’s settings without checking them.
Do I get charged for quotes or failed generations?
Quotes cost no generation credits, and a verified generation failure refunds once. Credits are reserved before submission. Queries and downloads also cost no generation credits, although reads and quotes remain subject to account request limits.
How do I switch my Seedance 2.0 integration to 2.5?
Change the model ID to seedance-2.5, refresh supported options through GET /models and quote the revised request. Keep the same request shape and asynchronous retrieval flow, and save signed outputs before they expire.
Try Seedance 2.5 Today
Ready to experience the next generation of AI video creation? Try Seedance 2.5 for free and see the difference for yourself.
