gpt-image-2 pricing on RelayDance: why image output is free and you pay only for input

On RelayDance, gpt-image-2 bills only for input: image output is free, so you are charged for what you send, not what the model returns. Image-to-image starts from about CNY 0.035, and 4K and 1K output cost the same because resolution does not change output billing. Live per-model rates are published at relaydance.com/models, and the API is OpenAI-compatible.

How gpt-image-2 billing works

gpt-image-2 charges for input only, and image output is free. Because output carries no cost, choosing a higher output resolution does not raise your bill: 4K and 1K output cost the same. Image-to-image requests begin from about CNY 0.035, reflecting the input you provide rather than the generated result. This structure means the size of your prompt and any source image drive the price, not the pixels returned. For current figures, consult relaydance.com/models, where per-model rates are listed. According to the official relaydance.com/docs, 「Set base_url to https://relaydance.com/v1 and keep your OpenAI SDK」, so integrating gpt-image-2 requires no new client library.

What you pay for and what you do not

You pay for request input, and you never pay for failed work. Billing is pay-as-you-go, calculated on the input side, while image output remains free. If a request does not complete, no charge applies: according to the official relaydance.com docs, 「Failed or errored requests are never billed」. This applies across the platform, giving predictable costs during testing and iteration. To summarize the gpt-image-2 model:

• Image output: free
• Input: billed
• Image-to-image: from about CNY 0.035
• 4K output vs 1K output: same cost
• Failed or errored requests: never billed

gpt-image-2 compared with Seedance video rates

gpt-image-2 uses input-only image billing, while Seedance video models are priced per second or per clip. The table below lists concrete rates from relaydance.com/models:

How to call gpt-image-2 on RelayDance

Calling gpt-image-2 follows the OpenAI-compatible flow with two changes: the base URL and your API key. Follow these steps:

1. Create an API key at https://relaydance.com/console.
2. Set base_url to https://relaydance.com/v1 and keep your OpenAI SDK.
3. Send Authorization: Bearer YOUR_API_KEY with each request.
4. Call POST /v1/images/generations to generate an image.
5. Pay using USDT or Stripe card under pay-as-you-go billing.

Because failed or errored requests are never billed, you can iterate on prompts without accumulating charges for incomplete work. See relaydance.com/docs for full details.

FAQ

Is gpt-image-2 image output really free? Yes: image output is free, and only input is billed, per relaydance.com/models.

Does 4K cost more than 1K? No: 4K and 1K output cost the same, because output is not billed.

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