General prompt
Natural visual language that you can read and revise.
AI Image Describer
Find the words for what you see.
Explore the subject, setting, light, and composition of an image in clear, editable language.
For a descriptive reading, choose General prompt and add “Describe the visible subjects, setting, composition, and lighting in plain language” to the optional direction.
From looking to making
A reference is only the beginning.
Capture the visual decisions that make an image work, then give them a direction of your own.
Read the whole image
Put the details into words.
Go beyond naming the subject. Build a useful description of composition, lighting, material, palette, and atmosphere.
Try your own reference →Built for control
Keep the details you love. Change the direction.
Separate the subject, composition, light, palette, material, and atmosphere. A structured reading makes it easier to preserve one decision while changing another.
Explore the output formats →Structured output
Clear labeled fields for repeatable workflows.
Model formats
Flux, Midjourney, Stable Diffusion, and JSON.
How to get a better result
01 / SOURCE
Use a clear reference
Choose an image where the subject and lighting can be read without heavy compression.
02 / PROFILE
Choose the depth
Concise is fast, Balanced is reusable, and Deep inspects smaller visual relationships.
03 / DIRECT
Add optional intent
Tell the system whether you care most about camera, material, layout, or atmosphere.
A little more clarity.
Answers before you upload.
What does the tool create?
It analyzes an image and returns a text prompt or structured description. You can edit the result and use it as a starting point in an image generator.
Which image files can I use?
Use a clear PNG, JPEG, or WebP file within the size limit displayed in the uploader. You can select a file, paste an image, or enter a public image URL.
Will it reproduce the original image exactly?
No. A prompt describes visible characteristics, but a different image model, seed, or setting will produce a different result. Review the description and adapt it to your purpose.
Can I choose a different prompt format?
Yes. The format control offers natural language and structured options, along with formats intended for common image model workflows.
Where can I learn to improve my prompts?
Our practical guides cover composition, lighting, prompt structure, and a first image-to-prompt workflow.