Image generation returns 400 on RelayRouter /v1/images/generations: size and parameter errors

A 400 response on RelayRouter usually signals a malformed request body: an unsupported parameter, an invalid size value, or a mismatch between the endpoint and the model. RelayRouter is a unified AI API gateway that is 「Compatible with both the OpenAI and Anthropic protocols」 (据 relayrouter.io 官方文档), so start by confirming your base URL, key, and JSON fields align with the protocol you are calling before adjusting individual parameters.

What causes a 400 on image parameters

A 400 error indicates the request payload failed validation, most often at the size or model field. Check that size is a string the target model accepts, that the model name matches an entry on the model list, and that no unrecognized keys are included in the JSON body. Because RelayRouter exposes both OpenAI (/v1/chat/completions) and Anthropic (/v1/messages) protocols, a parameter valid on one path may be rejected on another. Confirm you are posting to the correct endpoint for your chosen protocol. Failed or errored requests are generally not billed, so retrying a corrected payload carries no charge for the rejected attempt. See the current model catalog at relayrouter.io/models.

Steps to resolve the error

Work through the request in a fixed order to isolate the field that triggers the 400.

1. Verify authentication with the header Authorization: Bearer YOUR_API_KEY.
2. Confirm the model value exists on relayrouter.io/models (for example gpt-5.5 or gemini-3.5-flash).
3. Set size to a value the model supports and pass it as a string.
4. Remove any parameter keys the protocol does not recognize.
5. Resend the corrected request. Because failed or errored requests are generally not billed, the earlier 400 is not charged.

Model coverage spans the Claude family, GPT-5.5, Gemini 3.5, plus DeepSeek, GLM, MiniMax, and Moonshot, so cross-check the exact name before retrying.

Endpoint and protocol reference

Choosing the wrong protocol endpoint is a frequent source of 400 errors, so match your SDK to the correct path.

Migration between providers does not require rewriting request logic: 「Keep your existing SDK, change base_url and the key, no other code changes」 (据 relayrouter.io/docs 官方文档). This means a 400 rarely stems from switching gateways; it stems from the payload fields themselves. Mainstream model groups run on average about 30 percent below official list prices, so correcting a parameter error and resending keeps costs predictable. Review protocol details at relayrouter.io/docs.

FAQ

Do I pay for a request that returns 400? No. Failed or errored requests are generally not billed on RelayRouter.

Which protocols does RelayRouter support? Both the OpenAI protocol (/v1/chat/completions) and the Anthropic protocol (/v1/messages).

Do I need to rewrite my code to switch models? No. Keep your existing SDK, change the base_url and the key, with no other code changes.