THE FUTURE / AI REDESIGN BENCHMARK

Toilet paper
orientation.

How impressive can AI make one Wikipedia article? The same brief for every model. Explore the results and judge for yourself.

THE RESULTS

The leaderboard16

Highest score first · Every redesign opens in a new tab

Screenshot of the redesign by GPT 6 Astra Extra High #01
OpenAI

GPT 6 Astra Extra High

92/100
Model released

A distinctive editorial design with refined typography, a controllable 3D toilet roll and interactive comparisons that explain the subject.

Explore the redesign
Why this score?
Visual design
37/40
Originality
23/25
Interaction
18/20
Technical execution
14/15
Total
92/100

What impressesThe rotating roll, pullable paper, cat demonstration and history tabs form a coherent experience. Keyboard controls and source attribution show strong attention to detail.

Room to improveThe cat demonstration follows a predetermined animation, and the historical figures cannot be explored further. On mobile, the page is one pixel wider than the viewport.

Evaluated by AI on 14 Sept 2026· Release date source ↗

Screenshot of the redesign by GPT 6.1 Sol #02
OpenAI

GPT 6.1 Sol

89/100
Model released

An exceptionally polished interactive reinterpretation with strong typography, a rotatable 3D roll and clear content. Its combination of visual restraint and working experiments makes this one of the strongest redesigns.

Explore the redesign
Why this score?
Visual design
36/40
Originality
21/25
Interaction
18/20
Technical execution
14/15
Total
89/100

What impressesPulling, tearing, switching orientation, rotating and the cat test produce visibly different results. Arguments change by topic; the timeline, vote and expandable answers work. The original patent opens in a zoomable viewer. Source links, nuanced explanations and a polished mobile layout round out the experience.

Room to improveThe roll and cat test are scripted illustrations rather than a free physics simulation. The page heavily abridges the Wikipedia article, and its visual identity is less distinctive than the most original submissions. Reloading clears the vote; the site correctly states that this is not a public poll.

Evaluated by AI on 30 Sept 2026

Screenshot of the redesign by Claude Opus 5.5 #03
Anthropic

Claude Opus 5.5

88/100
Model released
Creation cost
$1.23
Tokens
131,900

A distinctive, cohesive design with large serif typography, detailed toilet rolls and substantial interaction tied to the subject.

Explore the redesign
Why this score?
Visual design
36/40
Originality
23/25
Interaction
18/20
Technical execution
11/15
Total
88/100

What impressesPulling, tearing and folding paper, weighing arguments, comparing two cat scenarios and taking a five-question quiz make the subject genuinely explorable. Your preference survives a reload.

Room to improveResetting the cat demonstration does not stop the roll: the counter starts increasing again after reset. The rolls are animated SVG illustrations without free 3D rotation; source attribution mostly points back to the original Wikipedia article.

Evaluated by AI on 23 Sept 2026· Release date source ↗

Screenshot of the redesign by Claude Sonnet 5.5 #04
Anthropic

Claude Sonnet 5.5

87/100
Model released
Creation cost
$0.97
Tokens
162,000

A convincing 3D bathroom with refined typography, an orbitable toilet roll and working paper interactions. Oversized article content on mobile lowers the score.

Explore the redesign
Why this score?
Visual design
36/40
Originality
23/25
Interaction
18/20
Technical execution
10/15
Total
87/100

What impressesPulling, tearing, both orientations, a fresh roll and the cat demonstration work. The timeline, local voting and decision helper are interactive; the helper applies its recommendation to the 3D roll. Votes and theme persist after reload.

Room to improveOn a 390-pixel mobile viewport, the article is 606 pixels wide, pushing text and components off-screen. The cat demonstration uses programmed random rules rather than a complete physics simulation.

Evaluated by AI on 28 Sept 2026

Screenshot of the redesign by MiMo V2.6 Pro #05
Xiaomi

MiMo V2.6 Pro

85/100
Model released
Creation cost
$0.07
Tokens
84,000

A polished editorial redesign with strong typography, an interactive SVG roll, an argument balance and a playful quiz with four outcomes.

Explore the redesign
Why this score?
Visual design
34/40
Originality
21/25
Interaction
17/20
Technical execution
13/15
Total
85/100

What impressesPulling and tearing paper, the hotel fold, patent annotations, weighing arguments and voting work. The quiz can be retaken, votes persist, and the page explicitly labels its poll as seeded demo data. Touch controls and reduced-motion mode also work.

Room to improveThe roll is a fairly flat SVG illustration without free 3D rotation. Source attribution contains no clickable references. Answering a quiz question with the keyboard drops focus back to the page rather than moving it to the next question.

Evaluated by AI on 24 Sept 2026· Release date source ↗

Screenshot of the redesign by GPT 6 Sol Extra High #06
OpenAI

GPT 6 Sol Extra High

84/100
Model released

A distinctive, polished editorial design with strong typography, a custom illustrated roll and clear structure. It works well on mobile, but its interactions are simpler than those of the leaders.

Explore the redesign
Why this score?
Visual design
35/40
Originality
21/25
Interaction
14/20
Technical execution
14/15
Total
84/100

What impressesThe roll responds to clicks and keyboard input. Argument cards and four scenarios change the explanation and orientation. A vote persists after reload and can be cleared. Source links and careful framing of historical figures show attention to detail.

Room to improveThe roll switches between two SVG drawings; rotation, pulling and tearing paper are absent. Scenarios display fixed explanations, while the timeline and figures are static. Pressing Escape in the mobile menu leaves keyboard focus on a hidden link.

Evaluated by AI on 23 Sept 2026· Release date source ↗

Screenshot of the redesign by meta.ai #07
Meta

meta.ai

82/100
Model released Unconfirmed

A polished interactive essay with distinctive typography, a pullable toilet roll, a horizontal timeline and a three-part personality test.

Explore the redesign
Why this score?
Visual design
35/40
Originality
22/25
Interaction
17/20
Technical execution
8/15
Total
82/100

What impressesThe roll changes orientation, paper can tear, and cat and toddler modes affect the result. The timeline, poll and quiz share a coherent editorial style.

Room to improveOn mobile, text and cards are clipped and dragging can remain stuck on PULLING. Some source details are misrepresented and source links are missing. The poll uses fixed starting counts and resets on reload.

Evaluated by AI on 14 Sept 2026· Exact model version and release date unconfirmed.

Screenshot of the redesign by GPT 6 Luna Max #08
OpenAI

GPT 6 Luna Max

81/100
Model released

A polished editorial design with strong typography, an illustrated toilet roll and an interactive timeline. The presentation is consistent and works well on mobile, but the interactive capabilities remain limited.

Explore the redesign
Why this score?
Visual design
35/40
Originality
20/25
Interaction
12/20
Technical execution
14/15
Total
81/100

What impressesThe over/under buttons and flip control change the illustration and save your preference. The five-event timeline works through buttons and a slider, including keyboard input. Historical figures are clearly distinguished from a current poll, and the source is linked.

Room to improveThe roll switches between two SVG views: free rotation, pulling and tearing paper are unavailable. Arguments and the percentage chart are static, with no quiz or simulation. After navigating through the mobile menu, its accessible button name incorrectly remains Close navigation even though the menu has closed.

Evaluated by AI on 23 Sept 2026· Release date source ↗

Screenshot of the redesign by Gemini 3.6 Thinking #09
Google

Gemini 3.6 Thinking

80/100
Model released

An extensive interactive workshop with a controllable 3D toilet roll, a cat demonstration, patent annotations and a step-by-step folding tutorial.

Explore the redesign
Why this score?
Visual design
33/40
Originality
22/25
Interaction
17/20
Technical execution
8/15
Total
80/100

What impressesPulling and tearing paper, comparing orientations, filtering arguments and practising a hotel fold give visitors several ways to explore the subject.

Room to improveThe mobile header is clipped. The global poll uses fixed starting figures and forgets your vote after reload. Some historical quotations cannot be traced to the cited sources.

Evaluated by AI on 14 Sept 2026· Release date source ↗

Screenshot of the redesign by DeepSeek V4.1 Flash #10
DeepSeek

DeepSeek V4.1 Flash

76/100
Model released

A complete, polished experience with strong typographic hierarchy, custom toilet-roll illustrations, a timeline and a working quiz.

Explore the redesign
Why this score?
Visual design
31/40
Originality
19/25
Interaction
15/20
Technical execution
11/15
Total
76/100

What impressesIllustrations, animated counters and scroll effects demonstrate that the model can go beyond a static restyling.

Room to improveInteraction remains simple: the quiz has one question and the charts cannot be explored. Some percentages cannot be traced to the Wikipedia article.

Evaluated by AI on 14 Sept 2026· Release date source ↗

Screenshot of the redesign by Claude Haiku 5.5 #11
Anthropic

Claude Haiku 5.5

74/100
Model released

A polished, playful redesign with a dark palette, large gradient typography, an interactive Canvas roll and a working quiz. Limited depth and keyboard problems lower the score.

Explore the redesign
Why this score?
Visual design
31/40
Originality
18/25
Interaction
16/20
Technical execution
9/15
Total
74/100

What impressesPulling and tearing paper, the slider, both orientations and O/U shortcuts work. Six argument cards flip; the four-question quiz has two outcomes and can be retaken. Voting preserves one changeable vote per browser and honestly discloses that limitation. The mobile page fits the viewport.

Room to improveThe roll is a simple 2D illustration with scripted animation. Argument cards cannot be reached by keyboard; the global Space shortcut tears paper instead of activating the focused button. Quiz answers lose focus. Source links, accessible selection states and reduced-motion support are missing; most history and source context are omitted.

Evaluated by AI on 8 Oct 2026· Release date source ↗

Screenshot of the redesign by Qwen 3.8 Omni Flash #12
Alibaba

Qwen 3.8 Omni Flash

74/100
Model released

A polished interactive guide with large serif type, an illustrated toilet roll, scenario advice and a sociology lab. Its many working features show creativity, but the simple simulation and several defects limit the finish.

Explore the redesign
Why this score?
Visual design
30/40
Originality
20/25
Interaction
16/20
Technical execution
8/15
Total
74/100

What impressesSix scenarios connect arguments to roll orientation. The reflection lab, keyword analysis and four-question quiz make the text explorable. Light/dark mode and the local demo poll persist after reload. Buttons work on mobile, and the roll supports keyboard controls.

Room to improveThe tile background sits over the roll and intercepts clicks and taps; the separate orientation buttons still work. Clicking the same vote repeatedly keeps increasing the count. The roll only switches between two drawings, text analysis uses fixed keywords, and clickable source references are missing. Some survey figures lack context, and the timeline contradicts the text about the decade of Oprah’s broadcast.

Evaluated by AI on 22 Sept 2026· Release date source ↗

Screenshot of the redesign by Claude Sonnet 5 #13
Anthropic

Claude Sonnet 5

72/100
Model released

A polished, readable guide with a calm palette, an illustrated toilet roll and a working personality quiz. The largely static presentation and small interaction defects limit its score.

Explore the redesign
Why this score?
Visual design
31/40
Originality
18/25
Interaction
13/20
Technical execution
10/15
Total
72/100

What impressesTypography, tiles, paper colors and illustrations form a coherent design. The over/under switch changes the roll and compares your choice with surveys. The four-question quiz has three outcomes, back navigation and retakes; explanatory text expands. Source attribution, the explicitly unscientific quiz and reduced-motion support show care.

Room to improveThe roll switches between two drawings; the timeline and chart offer no interactive exploration. The quiz stores only answer categories, so two different answers can appear selected at once. Choosing an answer loses keyboard focus. Expanded explanatory text is clipped when the viewport narrows until the panel is reopened.

Evaluated by AI on 14 Sept 2026· Release date source ↗

Screenshot of the redesign by Mistral Large 4 (Le Chonk) #14
Mistral AI

Mistral Large 4 (Le Chonk)

68/100
Model released

A polished, humorous magazine with warm paper colours, strong serif typography and working animations. Simple interactions, an incorrect roll illustration and weak source handling limit its score.

Explore the redesign
Why this score?
Visual design
31/40
Originality
19/25
Interaction
12/20
Technical execution
6/15
Total
68/100

What impressesCoherent chapters, drop caps, argument cards and a timeline give the subject a distinct editorial identity. Over/Under, spinning, O/U/S shortcuts, animated charts, voting, cat confetti and the closing flush animation work, including touch controls on mobile.

Room to improveIn Under, the paper sticks upward instead of hanging down behind the roll. Voting starts with invented counts and resets on reload without disclosing these limits. The 1965 counter-patent story and 1999 survey are unsupported; the page has no source links. At 390 pixels, the fixed bathroom wall extends 15 pixels beyond the viewport; reduced-motion handling and accessible selection states are missing.

Evaluated by AI on 6 Oct 2026· Release date source ↗

Screenshot of the redesign by Qwen 3.8 Flash #15

Interactions unavailable

Alibaba

Qwen 3.8 Flash

33/100
Model released
Creation cost
$0.03
Tokens
140,000

A polished opening screen with distinctive typography and a CSS toilet roll. A JavaScript error disables every custom interaction and keeps most of the redesign hidden.

Explore the redesign
Why this score?
Visual design
23/40
Originality
8/25
Interaction
1/20
Technical execution
1/15
Total
33/100

What impressesThe visible introduction combines a coherent paper palette, large serif type and a custom CSS toilet roll with shadows and gentle animation. The layout fits mobile, and ordinary navigation and source links work.

Room to improveLine 403 declares $ twice with const. The entire script fails to start: the roll simulation, poll, quiz, theme toggle and clickable roll do not work. All 37 reveal elements stay hidden, the quiz has no answer options and the canvas is blank. The score rewards only the visible result, not unreachable features described in the code.

Evaluated by AI on 21 Sept 2026· Release date source ↗

Screenshot of the benchmark error report for Mistral #16

Submission failed

Mistral AI

Mistral

0/100
Model released Unconfirmed

The supplied React code is incomplete and cannot compile. No working redesign is rendered.

View the error report
Why this score?
Visual design
0/40
Originality
0/25
Interaction
0/20
Technical execution
0/15
Total
0/100

What impressesThe code contains beginnings of a quiz, a poll and animations. There is no rendered result on which to award points.

Room to improveThe file ends halfway through a string: Unterminated string literal, line 479. The score applies to this incomplete file. The preview shows the benchmark error report; the original code is available.

Evaluated by AI on 14 Sept 2026· Exact model version and release date unconfirmed.

01 / THE BRIEF

One shared prompt.

Every model gets this Wikipedia article and the exact same brief. Complete freedom to show what it can do.

https://en.wikipedia.org/wiki/Toilet_paper_orientation
Create a beautifully designed, interactive version of this website, it must show your capabilities as a model. Make sure you do whatever you want, but redesign it to really impress me with your capabilities.

02 / THE EVALUATION

How impressive is it?

The score evaluates what the model demonstrates with this brief: design decisions, creativity and working interaction. Each submission is inspected in the browser, including on mobile.

Visual design
40points
Originality
25points
Interaction
20points
Technical execution
15points

A subjective AI assessment of this particular submission, not a general ranking of model intelligence. Highest total first; ties are ordered alphabetically. Submitted designs are preserved as supplied.