Share via


InputFidelity type

Control how much effort the model will exert to match the style and features, especially facial features, of input images. This parameter is only supported for gpt-image-1. Unsupported for gpt-image-1-mini. Supports high and low. Defaults to low.

type InputFidelity = "high" | "low"