Grok Imagine Image 2.0 Portrait Prompt Reference: Realistic and Artistic Dual Styles

A text-to-image prompt reference entry for developers, using the text-to-image pipeline of grok-imagine-image-2.0 to generate a portrait piece, for comparing the differences between two types of prompts—"realistic photography" and "artistic processing"—in terms of composition, lighting, and texture description. This piece uses a 3:4 vertical composition, 2K resolution, and medium quality tier. The subject is a single-person half-body portrait: the subject description focuses on facial structure, skin texture, and the direction of ambient light, while the style description layers on low-saturation film color grading and soft side backlighting, in order to observe the model's degree of fidelity to fine-grained portrait constraints. This schema accepts only four fields: prompt and quality / resolution / aspect_ratio, with no reference image input, so the piece relies entirely on text instructions for generation. It is suitable as a starting point for writing prompts for realistic portraits, concept character designs, and social media vertical avatar assets, and also makes it easy to batch-generate images with different aspect ratios under the same prompt structure.