ChartAnno - Implementation-Level Design Effectiveness: leaderboard

Metric: Design effectiveness (out of 5): mean of visual clarity, annotation organization quality and attention guidance, each an integer rubric score from 1 to 5 given by a GPT-5.4 judge, 0 for failed or missing output; annotated chart code generated from the unannotated code of 1,200 real-world charts; chart code input only; implementation-level instructions that specify the annotation elements and their visual properties. Source: arxiv.org. Saturation forecast: Around December 2026. 10 models tracked.

Top models

#ModelScore
1Gemini 3.1 Pro (Preview)4.21
2Gemini 3 Flash (Preview)4.13
3GPT-5.44.13
4Claude Sonnet 4.64.1
5Kimi K2.53.96
6Gemma 4 31B3.74
7Qwen 3.5 397B A17B3.74
8Qwen 3.5 122B A10B3.56
9Qwen 3.5 27B3.55
10Qwen 3.5 9B3.02

Interactive version: theaggregate.ai/benchmark?slug=chartanno-implementation-level-design-effectiveness · How It Works · Data refreshed daily, snapshot 2026-09-26.