Image to video API: how to animate an image with RelayDance

To animate an image with RelayDance, send a POST request to /v1/video/generations including your model, prompt, and seconds, then attach the source image in metadata.content[] and reference it in the prompt as @image1. RelayDance is OpenAI-compatible, so you set the base URL to https://relaydance.com/v1 and keep your existing SDK. Poll the task until status is succeeded, then read the video URL from the result.

How the image-to-video request works

You animate an image by placing it in metadata.content[] and citing it inside the prompt using @image1 to @imageN. A single request accepts up to 9 reference images, 3 reference videos, and 3 audio tracks, and clips can run up to 15 seconds. Additional metadata fields control ratio, resolution, generate_audio, and callback_url. According to the official relaydance.com/docs documentation, 「Set base_url to https://relaydance.com/v1 and keep your OpenAI SDK」. This means existing OpenAI client code can call RelayDance without rewriting the transport layer. Authentication uses an Authorization: Bearer YOUR_API_KEY header, with keys created in the console.

Steps to animate an image

Follow these numbered steps to turn a still image into a generated video clip.

1. Set base_url to https://relaydance.com/v1 and add your Authorization: Bearer YOUR_API_KEY header.
2. Upload or reference your source image and place it in metadata.content[].
3. Write a prompt that cites the image as @image1, and set seconds plus ratio, resolution, and optional generate_audio.
4. POST the payload to /v1/video/generations to create the task.
5. Poll GET /v1/video/generations/{task_id} until status is succeeded or failed, or set metadata.callback_url to receive the final state by webhook.
6. Read the video URL from the succeeded result.

Model pricing for image-to-video

Pricing is pay-as-you-go per generated video, with rates that vary by model and resolution. The table below lists concrete Seedance rates from relaydance.com/models. According to the official relaydance.com documentation, 「Failed or errored requests are never billed」, so retries after an error do not add cost.

For image models, gpt-image-2 bills only input (image output is free), image-to-image starts from about CNY 0.035, and 4K and 1K output cost the same.

Access, payments, and billing

Access is direct and OpenAI-compatible, with no BytePlus enterprise account or KYC required. You create API keys at the console and authenticate with a bearer token. Payments are supported through USDT and Stripe card, and billing is pay-as-you-go per generated video. Because failed or errored requests are never billed, error handling and retries stay cost-neutral. Live per-model rates are published at relaydance.com/models, and the API reference is at relaydance.com/docs. Available models include ByteDance Seedance 2.0 and Seedance 1.5 Pro, HappyHorse video models, and Grok Imagine video and image models, plus Grok text models.

FAQ

How do I reference the source image in the prompt? Place the image in metadata.content[] and cite it as @image1 to @imageN. A request supports up to 9 reference images.

How long can a generated clip be? Clips run up to 15 seconds, with duration set by the seconds parameter.

Am I charged if a request fails? No. According to the official relaydance.com documentation, 「Failed or errored requests are never billed」.