AI image descriptions are most useful when they are specific enough to explain the image but careful enough not to invent details. A good description names what is visible, explains how the frame is organized, and separates observation from interpretation.
This page focuses on examples. If you want the full method first, read How to Describe an Image for AI.
If you need the exact instruction to give an AI before it analyzes a photo, screenshot, chart, product, or portrait, use the AI image description prompts guide.
You can also upload a picture directly to the AI Image Describer when the tool is available and use the output as a draft to review.
A reusable image-description template
Use this order:
Summary: What is the main subject and setting?
Composition: Where is the subject, and how is the image framed?
Lighting and color: What direction, softness, contrast, and palette are visible?
Style or medium: Photograph, illustration, render, screenshot, diagram, or another format?
Important details: Materials, objects, readable text, patterns, or relationships that matter.
Uncertainty: What cannot be confirmed from the image?
You do not need every section for every image. The task decides how much detail matters.
Example 1: simple product photo
Image: A ceramic mug on a light wooden table.
Short description:
A matte white ceramic mug sits near the center of a light wooden table against a soft beige background. Diffused light from the left creates a gentle shadow to the right.
Detailed description:
Close product photograph of one matte white ceramic mug positioned slightly left of center on a pale oak tabletop. The handle faces right. A smooth warm-beige wall fills the background with no visible decorations. Broad soft light enters from the upper left, creating a gentle highlight on the mug and a soft contact shadow to the lower right. The palette is warm and neutral, with visible ceramic and wood texture.
Why it works: the description names the object, position, material, handle direction, background, light, and palette without adding a story.
Example 2: ecommerce shoe image
Three-quarter product photograph of a white running shoe with a pale gray sole, positioned on a light neutral studio surface. The toe points toward the lower-left corner while the heel sits farther back on the right. Soft overhead-left lighting reveals the mesh upper and creates a small contact shadow beneath the sole. No readable logo or text is visible.
If this description will become a generation prompt, continue with the Image-to-Prompt Generator.
Example 3: skincare bottle
Vertical product image of one amber-glass skincare bottle with a black dropper cap, centered on a pale stone block. The background is warm beige and softly graded. Broad side light from the left creates a long highlight on the glass and a soft shadow to the right. A cream rectangular label is visible, but the small text cannot be confirmed.
The phrase “cannot be confirmed” is more useful than guessing label text.
Example 4: professional portrait
Chest-up portrait of one adult person facing the camera with a slight natural smile. The subject wears a dark blazer over a light shirt and is centered against a plain light-gray background. Soft frontal-left light creates gentle facial modeling with minimal shadow. The crop leaves comfortable space above the head and around the shoulders.
Notice what is missing: no guessed name, profession, ethnicity, personality, or employer.
Example 5: casual outdoor portrait
Medium portrait of one person standing on a paved path outdoors, framed from approximately the waist up. The subject is placed on the right third of the image and looks toward the camera. Green foliage and a distant walkway are softly blurred behind them. Overcast daylight creates low contrast and natural skin tones.
Example 6: group photo
Wide group photograph of six people standing in two rows in front of a plain event backdrop. Three people stand in the back row and three in front. Most face the camera. The lighting is even and frontal, with minimal shadow. Several logos appear on the backdrop, but the text is too small to read confidently.
For people, describe visible arrangement and pose rather than guessing relationships between individuals.
Example 7: living room
Wide eye-level interior photograph of a modern living room. A cream sofa sits against the back wall, a low rectangular wooden coffee table is centered in the foreground, and a tall window on the left provides soft daylight. A floor lamp stands to the right of the sofa. The palette is cream, light brown, and muted gray, with moderate open floor space.
Example 8: kitchen interior
Wide photograph of a bright kitchen viewed from approximately eye level. White upper cabinets line the back wall above a pale stone countertop. A dark kitchen island occupies the foreground, and two simple stools stand on the near side. Daylight enters from a window on the right, producing soft natural shadows.
Example 9: real-estate exterior
Wide exterior photograph of a two-story light-colored house viewed from the front-left corner. A short driveway leads toward a closed garage on the right, while a small lawn and low shrubs occupy the foreground. The sky is bright with thin cloud cover, producing soft even daylight.
Avoid words like “luxury” unless the task specifically asks for an interpretive marketing description.
Example 10: food photograph
Close overhead photograph of a shallow white bowl containing pasta with red tomato sauce, grated cheese, and green herb leaves. The bowl sits on a dark wooden table with a folded beige napkin partly visible at the upper-right edge. Soft directional light from the left creates gentle texture and highlights on the sauce.
For prompt-writing around food, see the AI food photography prompts guide.
Example 11: coffee-shop scene
Medium-wide photograph of a small café interior. Two round wooden tables occupy the foreground, with dark chairs arranged around them. A service counter runs across the back of the room beneath warm pendant lights. Daylight enters from a large window on the left, mixing with warmer interior light.
Example 12: city street
Wide street photograph taken from sidewalk height. A row of mid-rise buildings extends into the distance on the left, while parked cars line the right side of the road. Several pedestrians are visible farther away. The sky is overcast, producing soft low-contrast light and muted colors.
Example 13: landscape
Wide landscape photograph of a calm lake surrounded by dark green hills. The horizon sits slightly above the vertical midpoint. A narrow wooden dock extends from the lower center into the water. Warm late-day light reflects near the horizon while the foreground remains cooler and darker.
Example 14: beach
Wide beach photograph with pale sand occupying the lower half of the frame and calm blue water across the middle. A single person appears small near the shoreline on the left. The sky is clear with soft haze near the horizon. Bright natural light creates short, faint shadows.
Example 15: pet photo
Close photograph of a golden-colored dog sitting on a wooden dock beside a lake. The dog faces the camera with its mouth open. Warm low-angle light highlights the fur, while tree-covered hills and still water form a softly focused background.
Do not infer temperament or emotion beyond what the image visibly supports.
Example 16: smartphone screenshot
Smartphone screenshot of a messaging application. A conversation occupies most of the screen with alternating message bubbles, while a text-entry field and send button appear along the bottom. The top bar contains a contact name and several interface icons. Individual message text should be transcribed only if clearly readable.
Example 17: analytics dashboard screenshot
Desktop screenshot of a web analytics dashboard. Summary metric cards appear across the top, including traffic and conversion indicators. A line chart occupies the central area, and a table of pages or sources appears below. A left sidebar contains navigation icons and labels. Exact values should be copied only when they are legible at the available resolution.
Example 18: code editor screenshot
Desktop screenshot of a dark-themed code editor. A file tree appears in the left sidebar, while the main pane shows source code with syntax highlighting and line numbers. A terminal panel is open along the bottom. The visible code appears to contain a function definition and several conditional statements, but exact text should be quoted only when clearly readable.
Example 19: chart
Bar chart comparing five categories along the horizontal axis. The bars increase from left to right, with the fourth category having the highest value. A vertical numerical scale appears on the left. The chart uses a simple light background and contains no visible legend.
When interpreting charts, separate what the graphic directly shows from any causal explanation.
Example 20: diagram
Simple process diagram with four rectangular boxes connected by arrows from left to right. Each box contains a short label, and the third step branches downward to a smaller decision box before rejoining the main flow. The diagram uses a white background with dark lines and blue accents.
Example 21: YouTube thumbnail
Landscape thumbnail composition with one large face on the left and a bright laptop screen on the right. The background is dark navy with strong blue highlights. Large high-contrast title text occupies the upper-right area. The composition is designed to remain readable at small size.
If you are creating rather than describing thumbnails, see the YouTube thumbnail prompts guide.
Example 22: social-media product image
Vertical social-media image showing one skincare bottle on a pale stone surface. The bottle sits on the lower-right third of the frame, leaving clean negative space in the upper-left area. Warm beige and cream dominate the palette, with soft side lighting and a subtle contact shadow.
Example 23: fashion photo
Full-length fashion photograph of one person standing against a plain concrete wall. The subject is centered, wearing a dark jacket and wide-leg trousers. Soft overcast outdoor light creates low contrast and visible fabric texture. The camera is approximately waist height with a straightforward perspective.
For generation prompts, continue with the fashion AI image prompts guide.
Example 24: logo or graphic mark
Simple black geometric symbol centered on a white background. The mark consists of two interlocking angular shapes with uniform stroke weight. No text is visible. The design uses flat solid color with no shading, texture, or three-dimensional effects.
Example 25: illustration
Flat digital illustration of a person sitting at a desk beside a large plant. Shapes are simplified with rounded edges, outlines are minimal, and the palette uses muted blue, green, and cream. The background is plain and the composition is centered.
Short vs detailed descriptions
A short description is useful when the task needs fast context:
A black keyboard sits on a pale desk beside a coiled white cable.
A detailed description is useful when composition or recreation matters:
Overhead product photograph of a compact black keyboard placed slightly left of center on a pale desk. A coiled white cable rests to the right, with the connector pointing toward the lower edge. Soft diffused light from the upper-left creates shallow shadows around the keys and cable. The palette is minimal: black, white, and warm off-white.
The longer version is not automatically better. It is better only when those extra details matter.
Description vs prompt
An image description explains what is visible.
A generation prompt explains what you want a model to create.
Description:
A red bicycle leans against a pale brick wall under soft afternoon light.
Prompt:
Red city bicycle leaning against a pale brick wall, three-quarter side view, soft afternoon light from the left, realistic street photography, warm neutral palette, no people, no text.
If you need the second format, use the Image-to-Prompt Generator or read How to Turn an Image Into an AI Prompt.
Common image-description mistakes
Guessing identity or intent. Describe visible facts first.
Inventing unreadable text. Mark text as unclear when you cannot verify it.
Listing objects without relationships. Explain where objects sit relative to each other.
Using only adjectives. “Beautiful, cinematic, professional” does not explain composition or light.
Over-describing irrelevant details. Match the description length to the task.
Ignoring uncertainty. A useful description can say what is unknown.
Frequently Asked Questions
What is an AI image description?
It is a text explanation of what appears in an image, often including the main subject, composition, lighting, color, medium, and important visible details.
How do I ask AI to describe a picture?
Upload the image and ask for the level of detail you need. A useful request is: “Describe the main subject, composition, lighting, colors, important details, readable text, and anything that is uncertain.”
What should an image description include?
Usually the main subject, visible action or setting, composition, lighting, dominant colors, medium or style, important details, and uncertainty when something cannot be verified.
Can an AI image description be used as alt text?
It can be a starting draft, but alt text should be adapted to the page context and reviewed for accuracy and relevance.
What is the difference between image description and image-to-prompt?
A description reports what is visible. An image-to-prompt workflow turns those observations into instructions for creating a new visual.