Chart Question Answering on chart-QA Internal curated
85.19AccuracyChartAgent
Evaluation results
| Method | Links | ||
|---|---|---|---|
| ChartAgent2025.10 | 85.19 | 3.42 | |
| GPT 52025.10 | 74.71 | 24.09 | |
| GPT 5-mini2025.10 | 73.18 | 11.24 | |
| Claude 3.7 Sonnet2025.10 | 69.71 | 15.52 | |
| GPT o4-mini2025.10 | 69.68 | 21.88 | |
| Gemini 2.02025.10 | 67.24 | 21.07 | |
| GPT-4.12025.10 | 66.61 | 24.32 | |
| GPT-o32025.10 | 62.93 | 9.14 | |
| Claude 3.5 Haiku2025.10 | 42.11 | 37.31 | |
| Mistral2025.10 | 38.54 | 38.74 | |
| o12025.10 | 33.07 | 44.31 | |
| GPT-4o2025.10 | 22.02 | 64.34 |