Put a new picture on the TV or a frame in a room photo
Swap what is on a TV screen or in a picture frame in a room photo by sending the room first and your artwork second to Ideogram 4.5 on Sume. Cost and checks.

To change what shows on a TV or in a frame, send the room photo as the first reference and your image as the second, and describe the swap in the prompt. In the Sume Image API, ideogram/ideogram-v4.5 edits the first reference and uses the others as sources, with up to four extra. A low-quality attempt costs $0.0375 on Sume, so you can try several framings.
The prompt
Be literal about both objects: "In the first image, replace the picture on the TV screen with the second image. Keep the room, the TV frame, the reflections and everything else unchanged." For a wall frame, swap in "the print inside the black frame above the sofa". Say "fills the screen edge to edge" if you want no borders; otherwise expect a margin.
Show the swap at the size people will see it. A TV that fills a third of a room photo is easy; one that is 5% of the frame gives the model few pixels to work with, so crop in, edit, and stitch the result back if the screen is small.
Use a straight-on photo if you can. A TV seen at a sharp angle gives the model less of the screen to work with, and the swapped picture is more likely to bend. If the room photo is yours to shoot, take one from the front of the screen and one from where the buyer will stand, and edit both with the same prompt.
Leave aspect_ratio out. The Sume docs say an Ideogram 4.5 edit without it keeps the source shape, so the room is not recropped.
Common failures
- The artwork is bent to fit a slanted screen. Ask for "correct perspective, matching the screen angle".
- A TV that is off shows a black glossy surface, and the model keeps some of that black. State that the screen is on and lit.
- Glass reflections vanish. If the original has a window reflection, add "keep the reflection of the window".
- The artwork's text is redrawn. Short text in the image can change, so check any words on your screenshot.
A tighter option for exact fit
If the corners must land on the screen, compute the warp yourself. Find the four corner points of the screen in pixels, warp your artwork with a perspective transform in Pillow or OpenCV, and paste it in. That gives exact pixels and no model. Use the model when you want glare, reflection and glow mixed in. You can combine them: paste first, then run a low-cost edit asking only to "blend the glare" over the pasted screen.
For a phone in a hand instead of a TV, the same two-image pattern is covered in putting an app screenshot into a phone photo.
Prices for this edit
One try per quality tier on Sume, from the provider list times 1.25. A typical session is three low tries to settle the wording and one medium final, which comes to $0.1875 in all.
| Quality | Per image | Use it for |
|---|---|---|
| low | $0.0375 | Placement and wording tests |
| medium | $0.075 | The default; a usual final |
| high | $0.275 | Close-up room shots with fine detail |
Sources
Related posts
More in Use cases
- Quiz feedback clips with an AI avatar: right and wrong answers
Render a short avatar clip per right and wrong quiz answer and let your LMS pick the one to play. Clip counts, cost by tier, and what Sume does not do.
- Recolor sneakers to sand suede, keep the laces white: prompt test
Four prompt wordings for a colorway edit that changes one part and keeps another, run on Ideogram 4.5 at the low tier on Sume for $0.15 in total.
- Remove a date stamp from a scanned photo: crop first, then a mask edit
Remove an orange date stamp from a scanned family photo: crop it for free in Pillow, or mask it and edit with openai/gpt-image-2.5 from $0.0094 an image.
- Remove a person or object from a finished clip with Gemini Omni edit
Delete a passer-by, a sign or a cable with one video_url request to Gemini Omni Flash 1.1. There is no mask field, so the prompt does the work. Cost per clip.
Written by Sume