Upload the photo
Choose a reference image that represents the look you want to describe.
PHOTO TO PROMPT
Use a photo as a visual reference and turn it into a structured image prompt you can edit and reuse.
Practical framework
Choose a reference image that represents the look you want to describe.
Focus on subject, composition, lighting, environment, colors, materials, and mood.
Treat the generated prompt as a starting point and adjust it to your creative goal.
PHOTO TO PROMPT EXPLAINED
A photo-to-prompt workflow turns a photograph into a written description that can be reused as an image-generation instruction. Instead of copying every visible detail by hand, you start with the photo and extract the visual choices that matter: subject, framing, camera perspective, lighting, environment, color, texture, and mood.
This is especially useful when a photograph has a composition or visual treatment you want to study and adapt. The goal is not to produce a magic sentence that recreates every pixel. The goal is to create an editable prompt that communicates the important visual structure and gives you control over what changes.
Identify who or what is shown, where the scene takes place, and the details that define the environment.
Describe framing, viewpoint, depth, lighting direction, shadows, focus, and the overall photographic feel when those details are visible and useful.
Capture the atmosphere, dominant colors, contrast, texture, and emotional tone that shape the photograph.
STEP BY STEP
A good result comes from deciding which photographic characteristics you want to preserve. Use the generated description as a draft, then adjust it to match your intended output.
Describe the subject before secondary details. For a portrait, this may include pose, clothing, expression, and visible features. For a product photo, it may include shape, material, finish, and placement.
Mention whether the shot is close-up, medium, or wide; where the subject sits in the frame; and how foreground, middle ground, and background are arranged.
Consider soft or hard light, direction, highlights, shadows, contrast, and whether the scene feels like studio lighting, daylight, evening light, or another visible setup.
Decide which characteristics should carry into the new image and which should change. This turns a description into a creative instruction instead of a simple caption.
EXAMPLE
Consider a photograph of a person sitting in a café near a window. A basic caption tells you what is in the image. A prompt can explain how the scene is photographed so the visual direction is reusable.
Basic caption
“A person sitting in a café by a window.”
Useful for identifying the scene, but it does not communicate the photographic treatment.
Prompt direction
“Candid lifestyle photograph of a person seated beside a large café window, soft natural side light, warm neutral interior, shallow depth of field, relaxed composition with subtle background blur, natural skin texture, understated editorial mood, eye-level perspective.”
This gives an image model more useful visual decisions while leaving room to change the subject or setting.
PRACTICAL TIPS
A photo can contain dozens of small objects. Keep details that affect the desired composition and remove distractions that do not matter.
Say whether the new image should feel like a product campaign, editorial photograph, cinematic scene, portrait, illustration, or another intended direction.
Use clear language that makes it easy to replace the subject, colors, setting, or mood without rewriting the whole prompt.
If the generated result misses an important visual feature, adjust one or two instructions at a time. This makes it easier to understand which change affected the result.
“Photo to Prompt” is commonly used when the reference is a photograph, while “Image to Prompt” is the broader term and can include illustrations, screenshots, artwork, and other visual references. If you are working with a wider range of references, visit Image to Prompt.
FAQ
They describe the same general workflow. Image to Prompt is the broader term.
Yes. The generated text is a starting point that you can edit.
Promptilot
Use the prompt tools to generate, refine, or expand your next instruction.