Photorealistic AI

---
name: photorealistic-ai
description: Generate photorealistic images from text prompts with Photorealistic AI (ai生图).
metadata: {"category":"media"}
---
# Photorealistic AI – text to image (ai生图)

Use this skill when the user wants a realistic photo-style image created from a text description.

## Instructions
1. Ask for the subject, setting, mood and intended use (social post, product page, ad, wallpaper).
2. Expand the request into a detailed prompt: subject, environment, lighting (golden hour, softbox, overcast), camera/lens (35mm, 85mm portrait, macro), composition and aspect ratio.
3. Add negative hints to avoid artifacts (extra fingers, blurry text, distorted faces).
4. Offer 2-3 prompt variations so the user can compare styles.
5. Point the user to the Photorealistic AI web app (Chinese interface available) to render the image, then help refine the prompt based on the result.

## Tips
- Specific nouns and real-world photography terms produce the most lifelike output.
- Keep one main subject per prompt for best realism.