AI kitchen counter product photo: scale cues for the right size
Image models often guess size. Give a real dimension in the prompt and include familiar objects of known size next to the product to keep scale believable.

To keep a product the right size in an AI kitchen counter photo, state the real dimension in the prompt and place familiar objects of known size beside it, such as a mug, a cutting board or a hand. Image models guess scale from context, and a blender that looks like a toy is a returns problem. Check the picture against a ruler.
Why scale drifts
A product reference shows shape and color but not size. When the scene is built around it, the model chooses a size that looks pleasant, not one that is true. Small items grow, large appliances shrink, and a shelf or countertop that does not fit the item makes the mistake visible.
Give it numbers and anchors
Put the height or width in the prompt and name two or three objects of known size. Say that the product should be about the height of a standard mug, or half the width of the cutting board.
curl -X POST https://api.sume.com/v1/images \
-H "Authorization: Bearer $SUME_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"google/nano-banana-pro","prompt":"Keep the product identical. Place this 30 cm tall stand mixer on a light oak kitchen counter, next to a standard 9 cm mug and a 40 cm wide cutting board, so the mixer is about three times the mug height. Morning window light, shallow depth of field.","aspect_ratio":"4:3","input_references":[{"type":"image_url","image_url":{"url":"https://example.com/mixer.png"}}]}'Measure the result
In an image editor, measure the product's height in pixels against the mug and compute the ratio. If your true ratio is 3.3 and the picture shows 2, ask for the size again with a stricter sentence or move the anchor next to the product. Perspective changes the numbers, so use a straight-on camera for checking.
When honesty matters most
For furniture, appliances and anything bought for fit, show the real size or do not show a scene at all. Keep one dimension drawing or a measured packshot in the listing. Treat the lifestyle image as mood, and do not let it replace the size information.
Variants without scale drift
For several scenes, repeat the same dimension and anchor sentence in each prompt, and re-run from the original packshot each time. See AI product photography prompts for a base template.
Related posts
More in Use cases
- AI music as the main focus of a Short: YouTube says to disclose it
YouTube lists music that is the main focus of a video as altered or synthetic content to disclose. How that applies to a Short built on a Sume music track.
- AI music video: a Lyria 3.5 track as the Timeline spine
Generate a song on the Music Router, then cut generated clips to it on Timeline 1.0. Track length is set in the prompt, not a duration field.
- AI name badge and business card art: art from AI, names from data
For badges and business cards, generate the design art with AI and set each person's name and contact lines from data. Ideogram 4.5 edits can fix one text line.
- AI nutrition label image: composite your real panel, don't generate it
A nutrition or ingredient panel is regulated text. Generate the package art, then place your real, proofread panel on it. Never let a model write the numbers.
Written by Sume