ChartMuseum: leaderboard
Chart-understanding benchmark reported in Anthropic system cards, with no-tools and Python-tools variants.
Metric: Overall Accuracy (%). Source: anthropic.com. Source: chartmuseum-leaderboard.github.io. Saturation forecast: Around January 2027. 22 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 80.7 | #58 |
| 2 | Gemini 3 Flash | 78.6 | #99 |
| 3 | GPT-5.2 (High) | 77 | #110 (GPT-5.2) |
| 4 | Gemma 4 31B (IT) | 71.1 | #176 |
| 5 | Gemma 4 26B A4B (IT) | 70.6 | #297 |
| 6 | GPT-5 Mini (High) | 63.3 | #186 (GPT-5 Mini) |
| 7 | Gemini 2.5 Pro | 63 | #152 |
| 8 | GPT-5 (High) | 62.9 | #98 (GPT-5) |
| 9 | O4 Mini (High) | 61.5 | #180 (O4 Mini) |
| 10 | O3 (High) | 60.9 | #124 (O3) |
| 11 | Claude Opus 4.5 (High) | 60.7 | #83 (Claude Opus 4.5) |
| 12 | Claude Opus 4.1 | 59.1 | #140 |
| 13 | Claude Sonnet 4 | 52.6 | #199 |
| 14 | Qwen 3 VL 30B A3B (Thinking) | 49.7 | #351 (Qwen 3 VL 30B A3B) |
| 15 | Qwen 3 VL 8B (Thinking) | 44.4 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=chartmuseum · How It Works · Data refreshed daily, snapshot 2026-10-20.