AI object placement
Place Objects, Furniture and People in Your Render with AI
Upload a render or photo, say what to add and where, and add a picture of the object if you have one. ArchiVinci generates a new image; with a picture, the editor is asked to fit the object to the space, perspective, scale and light of the scene.
Go to app and start rendering
See other modules below:
What is AI object placement in a render or photo?
AI object placement means adding something that a finished render or photo does not have yet: a piece of furniture, a plant, a lamp or people. In ArchiVinci's Edit module you upload the image, say in the prompt what to add and where, and add the object's picture if you have one.
ArchiVinci then generates a new image. You decide where the object goes; nothing plans a layout or measures the room.
How to Add Furniture or an Object to a Photo or Render
The steps are the same whether you add furniture to a photo or add an object to a photo or render of a room or building.
Open the Edit module with the button below and sign in or create an account.
Upload your render or photo under Input image: one image, JPG, PNG or WebP, up to 5 MB.
In Prompt, say what to add and where, for example a floor lamp to the left of the sofa. The prompt is always required.
Optional: pick tags under Add object (optional), such as Plants, Lamps or People. A tag adds its word to your request, never a picture.
Optional: for one specific piece, add its picture under Upload object (optional); for a specific spot, mark the area with the brush or the lasso under the image and confirm it.
Leave Render engine on the default and press Generate: one image costs one credit. The result opens in a before/after viewer, where you can download it.
Three Ways to Add an Object: Words, a Picture or a Marked Area
All three start from the same upload and need the prompt. Choose by how specific the object and the spot have to be.
Words Only: Decor, Plants and Lamps
Without a picture, the prompt does all the work: name the object, say what it looks like and where it stands relative to something in the scene. What comes back is drawn from your description, not one particular product; there is no object library to pick from.
The 14 tags under Add object (optional) add their words to the request: Car, Tree, Greenery, Sofa, Plants, People, Pet, Lamps, Accessories, Window, Coffee table, Paintings, Bed, Guitar.
A lamp added this way is an object in the scene. Changing the light of the image is a separate job.
This prompt names two objects and a place for each:
Add a black floor lamp to the left of the sofa and a tall potted plant next to the window.
A Picture of the Object: Your Furniture or Product in a Room Scene
For one specific piece, add its picture under Upload object (optional): one picture per run, up to 5 MB. It can be furniture you want to see in your own room or a product you make or sell. The prompt says where it goes.
With a picture, the editor is asked to fit the object to the space, perspective, scale and light of the scene. Use a picture that shows the whole object on its own.
For this example, a picture of an armchair was uploaded, so the prompt says "this armchair":
Add this armchair by the window, facing into the room.
In the example above, the armchair in the result follows the uploaded picture in shape, color and upholstery. A pattern, a logo or the proportions can differ, so compare the result with your picture.
A Marked Area: You Choose the Spot
To decide the spot yourself, mark the area with the brush or the lasso under the image, wide and tall enough for the whole object. Confirm the area and describe in the prompt what goes there. A mark lasts for one run: to repeat it, upload your image again and mark the same spot.
With a marked area, only that area changes: the rest of the picture is not redrawn. Whatever stands in the marked area can change or disappear, so mark clear floor or wall where you can.
For this example, one corner of a bedroom was marked, so the prompt says "here":
Place this armchair here, turned toward the bed.
How to Add People to a Render
You add people in words. Say how many people you want, what they are doing and where they are, and the editor adds them to the scene. Or pick the People tag and say in the prompt where they should be.
There is no people library and no way to choose who appears; every run generates new people. Check faces and hands before you use a result, and generate again if needed.
The example below used this prompt:
Add two people talking at the reception desk and one person sitting on the sofa with a laptop.
What stays the same when you add an object?
The editor is asked to keep everything else in the image, the composition and the camera angle. It is an instruction to an AI model, not a lock, so check small details.
On the default engine every result is a newly generated image, so details away from the new object can change too. Compare materials, patterns, lettering and whatever stands next to the new object with your original. To limit the change to one spot, mark the area first.
To change the light of a finished image, use AI relight; to see the scene from another camera position, see how to change the camera angle of a render.
Which tool do you need: add one object, furnish an empty room or change what is there?
Use this page when your render or photo is already composed and one thing is missing from it. Other jobs have their own pages:
If you want to | See |
|---|---|
fill an empty room with furniture from one description | |
change or replace what is already in a room photo | |
take furniture out of a photo | |
get a furnished view of an empty 2D floor plan, seen from above |
Frequently Asked Questions About Adding Objects with AI
Can I add furniture to a photo of my own home?
Yes. Upload the photo to the Edit module, say what to add and where, and add a picture of the piece if you have one. If you are preparing photos of a property for a real estate listing, see virtual staging.
Will the added object look like it belongs in the scene?
With a picture of the object, the editor is asked to fit it to the space, perspective, scale and light of the scene, but nothing guarantees the result. Check the object's size against a door and nearby furniture. If it is off, say the size in the prompt and generate again.
Can I add several objects at once?
In words, yes: one prompt can name more than one thing, as in the lamp and plant example. From pictures, it is one object per run, and a run without a marked area starts from your upload, not the last result. To build on a result, upload it or re-use it from the render history.
Can AI decide where my furniture should go?
No: nothing here plans a layout or measures a room. You say in the prompt where the object goes, or you mark the area. There is no control to drag, move, resize or turn the object afterward, so describe the new position and generate again.
Is it free to add an object to a photo with AI?
A new account gets 3 free credits. One image on the default engine, ArchiVinci 4.0, costs one credit, so that is three images. You need an account to upload and to generate.