Layout defect scenario · 2026-08-13

Detect broken responsive reflow in Smart home

Identify the single seeded visible layout defect.

Expected answerbroken responsive reflow

390 × 844 · 9 model configurations

Exact input

Prompt

You are evaluating a user interface screenshot for objective rendering defects.

Check only for these visible issue categories: viewport_cropping, container_clipping, text_overflow, element_overlap, obscured_actionable_control, or broken_responsive_reflow. Do not report subjective design preferences, copywriting concerns, or accessibility assumptions.

If there is one visible issue, return its category and the tight affected region on a normalized 0 to 1000 grid. If none of these issues is visible, return has_issue=false and null category and region values.

Return only data matching the supplied JSON schema. Do not describe your reasoning.
SHA-256 · 7d8cfe281f4ef9b625bbcc6c89c45236b392624e96c23044e6dcaac1d8d7adb0

Ground truth

Smart home layout-defect benchmark screenshotExpected defect
Ground truth: broken responsive reflow

Model evidence

9 model results

Blue marks the seeded defect region. Orange marks the model’s returned region when available.

Holo3 122B A10B

H Company · holo3-122b-a10b
Incorrect
Latency
1.17 s
Category
Localization IoU
0.000
Cost
$0.00025
Attempts
1
                  {"has_issue":false,"category":null,"region":null,"description": "No issues found."}
                

Grok 4.5

xAI · x-ai/grok-4.5
Correct
Latency
5.51 s
Category
broken responsive reflow
Localization IoU
0.691
Cost
$0.0016
Attempts
1
                  {"has_issue": true, "category": "broken_responsive_reflow", "region": {"x": 80, "y": 430, "width": 840, "height": 280}, "description": "Quick view list items misaligned; card 2 is indented and narrower than cards 1 and 3, breaking consistent layout."}
                

GPT-5.4

OpenAI · openai/gpt-5.4
Incorrect
Latency
9.30 s
Category
Localization IoU
0.000
Cost
$0.0033
Attempts
1
                  {"has_issue":false,"category":null,"region":null,"description":""}
                

Gemini 3.6 Flash

Google AI Studio · google/gemini-3.6-flash
Correct
Latency
9.58 s
Category
container clipping
Localization IoU
0.000
Cost
$0.0044
Attempts
1
                  {"has_issue":true,"category":"container_clipping","region":{"x":355,"y":510,"width":599,"height":70},"description":"The card for 'Kitchen lights' is displaced horizontally and clipped at the right edge."}
                

GPT-5.6 Sol

OpenAI · openai/gpt-5.6-sol
Correct
Latency
11.32 s
Category
viewport cropping
Localization IoU
0.000
Cost
$0.0122
Attempts
1
                  {"has_issue":true,"category":"viewport_cropping","region":{"x":354,"y":501,"width":646,"height":67},"description":"The Kitchen lights card extends beyond the right edge of the viewport and is visibly cropped."}
                

GPT-5.6 Luna

OpenAI · openai/gpt-5.6-luna
Correct
Latency
9.13 s
Category
container clipping
Localization IoU
0.000
Cost
$0.00018
Attempts
1
                  {"has_issue":true,"category":"container_clipping","region":{"x":354,"y":510,"width":646,"height":69},"description":"The second quick-view card is shifted to the right and clipped by the viewport edge."}
                

Claude Sonnet 5

Anthropic · anthropic/claude-sonnet-5
Incorrect
Latency
10.87 s
Category
broken responsive reflow
Localization IoU
0.000
Cost
$0.0037
Attempts
1
                  {"has_issue":true,"category":"broken_responsive_reflow","region":{"x":141,"y":432,"width":249,"height":66},"description":"The 'Kitchen lights' quick view card is misaligned and shifted right, overlapping/extending beyond the container edge instead of stacking properly with the other cards."}
                

/description must NOT have more than 160 characters

Claude Opus 5

Anthropic · anthropic/claude-opus-5
Incorrect
Latency
10.92 s
Category
viewport cropping
Localization IoU
0.000
Cost
$0.0092
Attempts
1
                  {"has_issue": true, "category": "viewport_cropping", "region": {"x": 340, "y": 1055, "width": 660, "height": 100}, "description": "The 'Kitchen lights' list row is shifted right and extends past the right viewport edge, cutting off its content and chevron."}
                

region must be contained in normalized 0..1000 coordinates. has_issue=true requires a category and normalized region.

Gemini 3.1 Pro Preview

Google AI Studio · google/gemini-3.1-pro-preview
Correct
Latency
10.05 s
Category
broken responsive reflow
Localization IoU
0.164
Cost
$0.0034
Attempts
1
                  {"has_issue": true, "category": "broken_responsive_reflow", "region": {"x": 355, "y": 510, "width": 645, "height": 69}, "description": "The 'Kitchen lights' card is severely misaligned and shifted to the right, breaking the layout grid."}
                
EN