Which video model to use: Seedance 2.0 vs 1.5 Pro vs HappyHorse vs Grok Imagine

Choose by output goal and budget: Seedance 2.0 covers standard production at about $0.190 per second (720p) or $0.470 per second (1080p), Seedance Fast runs at about $0.152 per second for quicker drafts, and native 4K is priced at about $4.90 per 5 second clip. Seedance 1.5 Pro, HappyHorse, and Grok Imagine are also available through one OpenAI-compatible API at relaydance.com/models. Live per-model rates are published there.

How Seedance 2.0 pricing compares across resolutions

Seedance 2.0 is billed per second, with the rate rising as resolution increases. According to relaydance.com/models, 720p output is about $0.190 per second and 1080p output is about $0.470 per second, while Seedance Fast is about $0.152 per second for lower-cost generation. Native 4K is priced differently, at about $4.90 per 5 second clip. Because billing is pay-as-you-go per generated video, you pay only for what completes. According to the official relaydance.com docs,"failed or errored requests are never billed", so retries during testing do not add cost. Use these numbers to estimate spend before submitting tasks.

How to choose between the four model families

Match the model to the task: standard 720p or 1080p clips, fast drafts, native 4K, or image generation.

Seedance 1.5 Pro, HappyHorse, and Grok Imagine video and image models are available through the same endpoint; check live rates at relaydance.com/models.

How to submit a request on the OpenAI-compatible API

All models use one OpenAI-compatible interface, so you keep your existing SDK and change one setting. According to the official relaydance.com/docs,"set base_url to https://relaydance.com/v1 and keep your OpenAI SDK".

1. Create an API key at the console and set Authorization: Bearer YOUR_API_KEY.
2. Set base_url to https://relaydance.com/v1.
3. Submit a video task: POST /v1/video/generations with model, prompt, and seconds.
4. Add metadata (ratio, resolution, generate_audio, callback_url, content[]).
5. Poll GET /v1/video/generations/{task_id} until status is succeeded or failed, then read the video url.

You can attach up to 9 reference images, 3 reference videos, and 3 audio tracks per request, with clips up to 15 seconds. See relaydance.com/docs for details.

How image generation is billed with gpt-image-2

Image billing differs from video: with gpt-image-2 the image output is free and you are charged only on input. According to relaydance.com/models, image-to-image starts at about ¥0.035, and 4K output is priced the same as 1K output. This means resolution does not change the per-image output cost for gpt-image-2. Image requests use POST /v1/images/generations on the same base URL. As with video, failed or errored requests are never billed, so test prompts without accruing charges. Payments are accepted via USDT and Stripe card, and access does not require a BytePlus enterprise account or KYC.

FAQ

Are failed requests charged? No. According to the official relaydance.com docs,"failed or errored requests are never billed".

Do I need a new SDK? No. Keep the OpenAI SDK and set base_url to https://relaydance.com/v1.

How many reference media can I attach? Up to 9 reference images, 3 reference videos, and 3 audio tracks per request, with clips up to 15 seconds.