RankDude

How I test AI interior design apps

Ten tests, the same for every app, each graded from 1 to 5 on a fixed scale. The overall grade is their average plus my verdict, which can move it by up to one point. Every test is run by hand on an account I use, and the photos, renders, overlays, timings and screenshots behind each grade go in the full test results published with the review.

Written by Miles Foster, Owner and tester, RankDude.

Each app gets the same photos of real rooms: an empty living room, a furnished bedroom, a dated kitchen, a small bathroom, a home office and a studio flat, plus a house front, a patio and a measured floor plan of a two-bedroom flat. Every room is redesigned in the same three styles (Scandinavian, industrial and mid-century) and from the same written prompt naming six specific items. Nothing is graded from a showcase or a press kit.

Each review states how the account was obtained: bought at the public price, used on a free tier, or provided by the vendor. A vendor-provided account never changes a grade, and a vendor never sees a review before it is published.

How I rate

Every tool gets one score out of 5.

First, I run the same 10 tests on every tool and give each one a grade from 1 to 5. The average of the graded tests is the starting point; a test that cannot be graded is left out of it, and the review says how many tests the score is based on.

Then I add my verdict: what the tests cannot measure, such as how the tool holds up after the test weeks, how billing and cancelling work in practice, and how it compares with the others in the category. My verdict can move the score up or down by one point at most.

Every review shows how much my verdict changed the score, and why, in its scores table.

What a grade means

5Excellent
4Good
3Average
2Below average
1Poor

The final score is the average of the graded tests (ten when every test can be graded), to one decimal place, plus my verdict, which can move it by up to one point. If a test couldn't be finished properly it gets no grade, and the review says how many tests the score is based on. Some tests on a free plan use fewer renders than usual because of credit limits; I test them again on a paid plan when I can.

The 10 tests

1.Render Realism

Each of the six rooms is rendered in the three styles and the first result is kept. All 18 are inspected at 100% for bent straight lines, floating or melted furniture, wrong scale, impossible shadows or reflections, and garbled patterns or text. A render is clean when none of these shows.

  • 5: 16 to 18 clean
  • 4: 13 to 15
  • 3: 9 to 12
  • 2: 5 to 8
  • 1: 4 or fewer

2.Fidelity to the Real Room

Each render is laid over its original photo at 50% opacity. It counts as faithful only when walls, windows, doors, ceiling height and fixed fittings (radiators, sinks, sockets, beams) are all where they really are and the same size. Moved or invented windows and doors are counted separately.

  • 5: 17 or 18 faithful
  • 4: 14 to 16
  • 3: 10 to 13
  • 2: 5 to 9
  • 1: 4 or fewer

3.Style Control

Each style has a short list of defining traits written beforehand, and the written prompt names six specific items (for example, a green velvet sofa and an oak floor). Every render is checked for the traits and the items. The share present across all rooms is the grade.

  • 5: 90% or more present
  • 4: 75% to 89%
  • 3: 55% to 74%
  • 2: 30% to 54%
  • 1: under 30%, or styles look the same

4.Targeted Edits & Staging

On the living room and the kitchen I repaint the walls to a given colour, change the floor to light oak, swap only the sofa or the table, remove one object, and furnish the empty room from scratch. Each of the ten edits counts when the change is made and the rest of the image is unchanged side by side.

  • 5: 9 or 10 correct
  • 4: 7 or 8
  • 3: 5 or 6
  • 2: 3 or 4
  • 1: 2 or fewer, or no targeted edits

5.Furniture You Can Buy

I follow 20 product links taken from the renders. A link counts when the page is live, the item can be bought and delivered to my country, and it matches the item shown in colour, shape and rough size. The price shown in the app is compared with the price at the shop.

  • 5: 16 to 20 match
  • 4: 12 to 15
  • 3: 7 to 11
  • 2: 1 to 6
  • 1: none, or no product links

6.Speed per Render

Each of the 18 style renders is timed from pressing generate until the full-quality result is shown; progress screens and previews do not count. The median is the grade; the average, fastest, slowest and failed renders are reported.

  • 5: median under 15 seconds
  • 4: 15 to 29 seconds
  • 3: 30 to 59 seconds
  • 2: 60 to 179 seconds
  • 1: 3 minutes or longer

7.Room & Space Coverage

Living room, bedroom, kitchen, bathroom, home office, child's room, house front, garden or patio, a commercial space and an empty room for staging are each tried with the matching photo or a stand-in. A space counts when the app supports it and the render passes the realism and fidelity checks.

  • 5: all ten
  • 4: 8 or 9
  • 3: 6 or 7
  • 2: 4 or 5
  • 1: 3 or fewer

8.3D & Floor Plans

The floor plan of the two-bedroom flat is uploaded. I check six points: the plan is read correctly, a 3D model is built, it can be walked through, furniture is placed to scale, the plan can be exported, and five dimensions I measured with a tape are within 5% in the app.

  • 5: all six, dimensions within 5%
  • 4: five
  • 3: three or four
  • 2: one or two
  • 1: no plan or 3D features

9.Ease of Use

From opening the app with a new photo of the bedroom, I count the steps and time to the first restyled render, then to swapping the bed for another. Each tap, menu, upload and dialog counts as a step; help pages and dead ends are noted.

  • 5: first render under 2 minutes and 6 steps
  • 4: under 4 minutes and 10 steps
  • 3: under 8 minutes, or help needed once
  • 2: 8 minutes or longer
  • 1: a task could not be finished

10.Pricing & Value

Every plan and credit pack is read from the checkout, not the marketing page. The cost of one render at the plan's full resolution with no watermark is worked out in dollars, and the long edge of the downloads is measured in pixels on every plan. Free-tier limits, renewal terms and payment methods are listed as seen at checkout.

  • 5: under $0.10 per render, or unlimited, at 2,048 pixels or more
  • 4: $0.10 to $0.24
  • 3: $0.25 to $0.49
  • 2: $0.50 or more, or watermarks or low resolution on paid plans
  • 1: poor value or hidden costs

Documented but not graded

Video walk-throughs, augmented reality previews on a phone, mood boards, sharing with a designer or contractor and cost estimates for a renovation are checked and described in every review. When they work well they count in the written assessment, but they never add points to the overall grade.

Rules that apply to every test

  • Real use of the product on a real account, never promotional descriptions or showcase galleries.
  • Identical photos, styles, prompts, tests, thresholds and sample sizes for every app in the category.
  • Authorised accounts only; no password sharing, no bypassing of credit limits, watermarks or bot checks, and only photos of homes I may photograph.
  • Every grade has evidence: the original and rendered images, an overlay, a measurement, a screenshot or a table, in the full test results.
  • A grade changes only when a retest produces new evidence; the review records the date and the reason.

Questions about a specific grade go to the contact page. Makers can submit an AI interior design app for testing. No AI interior design app has been through these tests yet; every one I follow is listed on the AI interior design apps page, and reviews appear there as tests finish.