XTC-Bench - Attribute Query: leaderboard

XTC-Bench (2026) attribute questions on COCO and Visual Genome images, such as an object's colour or material, judged against the scene-graph reference. Score 0-1, shown times 100; higher is better.

Metric: Understanding score (0-1, times 100) on attribute questions about objects, a Qwen3-235B judge scores each fact 0-5, normalized to 0-1 and shown times 100, over XTC-Bench's 2,000 scene-graph-annotated images from COCO 2017 val and Visual Genome (over 31,000 atomic facts about objects, attributes and relations); higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 10 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Gemini 2.5 Flash Image70.8
2BAGEL-7B-MoT70.4#4619
3GPT-570#178
4BLIP3-o8B67.7
5OmniGen266.6#4579
6Janus-Pro-7B66.5#6857
7Show-o2-7B66.4
8Tar-7B59.1
9Show-o57.9
10MMaDA-8B48.4

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=xtc-bench-attribute-query · How It Works · Data refreshed daily, snapshot 2026-11-21.