Documentation

Everything you need to get reliable outputs from both AI image tools.

Use this documentation to move from first upload to consistent results. It covers prompt generation, image description, mode selection, and practical quality improvements for real creative work.

1. Choose the right tool

Image to Prompt is best when you need a generation-ready prompt. AI Image Describer is best when you need analysis, extraction, or a direct answer about the image.

2. Pick the correct mode

Use general or structured modes for broad prompt drafting, Midjourney or Stable Diffusion modes for platform-specific phrasing, and object/text/art-style modes for targeted analysis.

3. Refine for your workflow

Treat the first result as a strong draft. Add constraints, remove noise, and adjust wording for your model, brand, or visual direction.

Prompt modes

  • General: a solid all-purpose prompt draft.
  • Structured: labeled sections for easier editing and reuse.
  • Graphic: more design-system and layout oriented outputs.
  • JSON: machine-readable structure for downstream workflows.
  • Flux / Midjourney / Stable Diffusion: tuned for platform-specific prompt patterns.

Description modes

  • Detail / Brief: long-form or short-form summary.
  • Person / Objects: focused subject breakdowns.
  • Art Style: visual style analysis for references and moodboards.
  • Extract Text: OCR-style text transcription from visible image content.
  • Custom: answer a user question about what is visible.

Best practices for better outputs

  • Start with a clear, well-lit image whenever possible.
  • Choose the mode that matches your end goal instead of always using general mode.
  • For prompts, add missing constraints like lens, atmosphere, aspect ratio, or negative prompts after generation.
  • For descriptions, ask a focused custom question when you need only one decision-critical answer.

Troubleshooting

  • If the output is too generic, try a more specific mode like structured or art style.
  • If text extraction misses details, upload a sharper crop or higher-contrast image.
  • If a prompt is too long, trim repeated adjectives and keep the most important visual anchors.
  • If results feel off, verify that the uploaded file matches the actual creative reference you intended.

Need examples and prompt-writing guidance?

Browse the blog for workflow examples, comparison articles, and prompt-improvement strategies based on real image references.