To get the best results, please follow these guidelines when uploading files and writing prompts:

1. About your uploaded file

• Make sure the image or file is clear and complete (not heavily blurred, cropped, or corrupted).

• Common formats such as JPG, PNG, and WEBP work best.

• If your image contains people, objects, or backgrounds you want changed, describe what can change and what does not need to be preserved.

2. How to write a good prompt

• Be clear about the goal of the edit, not just the style.
  (e.g. “create a fictional illustration based on the pose” instead of “make it look exactly the same”)

• Avoid asking the AI to identify, confirm, or preserve real-world identities.

• If likeness is not important, explicitly say so. This greatly improves success rates.

• For background-related requests, specify whether the background should be kept, edited, or removed.

3. If the AI says it “cannot process”
This does not mean your file is broken. Common reasons include:

• The request tries to preserve or recreate a real person’s identity.

• The prompt is ambiguous or contains conflicting instructions.

• The requested transformation goes beyond safe or supported editing boundaries.

4. How to fix a failed prompt

• Rephrase your request to focus on fictional, stylized, or abstract results.

• Remove phrases like “keep the same person” or “make it look exactly like the original”.

• Clarify that the output does not need to resemble any real individual.

• Simplify the prompt: describe what you want to see, not how the AI should decide internally.

💡 Tip:
When in doubt, treat your input as a reference rather than something that must be exactly preserved. This gives the AI more freedom and leads to better results.