[1] Generated with the F.1 base model, other portrait F.1 large models can also be tried.

[2] Standard sampling steps around 30, prompt coefficient 2.5-3.5, sampler Euler,

[3] Output image dimensions: 896*1280, 1024*1024, 1024*1536, 1536*2048

[4] Prompts can include: Photo of a young woman standing in (a certain scene), [clothing], [action], [environment], [atmosphere], etc.; default negative prompts are sufficient. Primarily input in natural language, but additional phrases can also be used to assist the prompt.

[5] This model is suitable for realistic human photography scenarios, achieving good realistic effects in both indoor and outdoor settings.