How to redesign a room from one photo

Step 1 — Start with a photo that shows the room, not a corner of it

Stand in the doorway, get as much of the room in frame as your phone allows, and shoot in daytime light if you can. You are not trying to make the room look good — you are trying to give the tool an accurate read on proportions, existing furniture, and where the light comes from. A photo of one accent chair tells it almost nothing; a photo of the whole room tells it what it is actually working with.

If you are unsure what counts as a usable photo, the full photo guide covers light, angle, and framing in more detail. The short version: natural light, whole room in frame, nothing important blocked.

Step 2 — Decide what you're actually testing before you pick a style

This is the step people skip, and it is the reason a lot of first attempts feel unsatisfying. "Redesign my living room" is not a decision — it is a vague direction. Before you upload anything, get specific about what you are trying to learn:

• Are you testing a color or mood change, or a full furniture swap?
• Is the current layout the problem, or is it the palette?
• Are you deciding between two directions you already have in mind, or genuinely exploring?

If you are not sure whether the fix is paint, rearranging what you have, or replacing a piece entirely, this decision breakdown is worth reading first — it changes what you should even be testing with a redesign tool.

Step 3 — Upload, pick a direction, and generate

Upload the photo — you'll sign in first, since every step here runs through your account, though uploading and the free photo analysis don't spend a credit. Choose one of the suggested directions that best matches what you decided in step two. The first direction step does not accept a free-text note. If you want to change a color, material, or general furniture idea, generate the first result and use the refinement box described in step five. Generating is the step that actually spends credits, and you will see the exact cost before you confirm.

Generation typically takes one to two minutes per direction. That is a real range, not a marketing number — see how it works for the full timing breakdown and what happens if it runs long.

Step 4 — Read the result like evidence, not a verdict

When the result comes back, your original photo sits next to it so you can compare directly. Resist the urge to treat the first result as final. Instead, ask three questions:

• Does the mood match what I was actually going for? If not, that is more useful information than a "correct" result — it narrows down what you want.
• Is there one element doing most of the work? Often it is the wall color or the flooring tone, not the furniture. That is a cheap thing to test in real life before buying anything.
• What would I need to verify before spending money on this? Furniture fit, wall color under your actual lighting, and real prices are never confirmed by the image itself. See limits & privacy for the full boundary list.

Step 5 — Iterate on one thing at a time

If a direction is close but not right, do not start over. Refine the specific part that is off — the wall color, the furniture arrangement, one piece — instead of regenerating the whole room from scratch. Each small change is easier to evaluate than a whole new image, and it keeps your comparison anchored to the direction you already liked most.

If nothing in a style is landing, try a genuinely different direction rather than nudging the same one repeatedly. Two or three distinct directions usually show you the real range faster than five variations of one.

A worked example: one photo, three directions

It helps to see this play out concretely. Take a single, fairly ordinary living room — plain walls, decent daylight, nothing staged. Run it through three different directions and the differences tell you more than any one of them alone:

• Warm minimal keeps the existing floor tone and window light, and just quiets the palette down — soft neutrals, fewer competing textures.
• Scandinavian pushes lighter wood tones and a lower furniture profile, which reads differently in the same light.
• Moody contemporary goes the other way — deeper wall color, layered textiles, warm low lighting.

Looking at all three against the same original photo usually makes one direction obviously more "you" than the other two, faster than describing your taste in words would. The full walkthrough of this exact comparison, including what to double-check before buying anything from each direction, is on the examples page.

Common mistakes worth avoiding

• Uploading a dark or heavily cluttered photo and expecting a clean result — the tool works with what it can see, and a hidden room stays hidden in the output.
• Expecting exact furniture matches. A result shows a style and a mood, not a shoppable catalog of the specific pieces pictured.
• Treating a result as a quote. Nothing in a redesign direction implies a real-world price for materials, furniture, or labor.
• Generating five near-identical variations instead of testing genuinely different directions. More attempts only help if they are asking a different question.
• Skipping the decision step. Uploading a photo before deciding whether you're testing paint, layout, or furniture usually means picking a style that doesn't actually answer your real question.

Most of these come down to the same root cause: treating the tool as the decision-maker instead of a fast way to see options you've already narrowed down. Do the narrowing first, and the results get a lot more useful.

Quick questions