LLM Stats (ChartQA): leaderboard
Visual QA on bar charts, line graphs, and pie charts from real-world sources, requiring reading values, identifying trends, and performing arithmetic over chart data.
Metric: Score (%). Source: llm-stats.com. Saturation forecast: Estimated already saturated. 26 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude 3.5 Sonnet | 90.8 | #350 |
| 2 | Llama 4 Maverick | 90 | #464 |
| 3 | Qwen 2.5 VL 72B Instruct | 89.5 | #377 |
| 4 | Nova Pro | 89.2 | #569 |
| 5 | Llama 4 Scout | 88.8 | #656 |
| 6 | Mistral Small 3.2 | 87.4 | #605 |
| 7 | Qwen 2.5 VL 7B Instruct | 87.3 | #654 |
| 8 | Nova Lite | 86.8 | #754 |
| 9 | GPT-4o | 85.7 | #343 |
| 10 | Llama 3.2 90B Instruct | 85.5 | #601 |
| 11 | Qwen2.5-Omni-7B | 85.3 | #624 |
| 12 | Llama 3.2 11B Instruct | 83.4 | #1126 |
| 13 | Pixtral-12B | 81.8 | #808 |
| 14 | Phi-4 Multimodal Instruct | 81.4 | #906 |
| 15 | Gemma 3 27B | 78 | #607 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=llm-stats-chartqa · How It Works · Data refreshed daily, snapshot 2026-10-20.