ChartArena - Combination Chart (English): leaderboard
Metric: mAP-high (%) on English combination charts, each model parses the chart image into its native structured format, normalized format-agnostically to (header, entity, value) triples for numeric charts or to labeled graphs and trees for flowcharts and mind maps; mean average precision over similarity thresholds at the high tolerance level, averaged over three visual styles (digital renderings, printed photos, hand-drawn photos); higher is better. Source: arxiv.org. Saturation forecast: Around July 2027. 44 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 39.7 |
| 2 | Kimi K2.5 (Non-reasoning) | 33.6 |
| 3 | Seed 2.0 Pro (Non-reasoning) | 32.4 |
| 4 | Qwen 3.5 35B A3B (Thinking) | 31.5 |
| 5 | Qwen 3 VL 235B A22B Instruct | 29.1 |
| 6 | Gemini 2.5 Pro | 28.7 |
| 7 | Gemma 4 31B (IT) | 22.7 |
| 8 | GLM-4.5V | 21.2 |
| 9 | MiMo-V2-Omni | 19.4 |
| 10 | Qwen 3.5 9B | 16.8 |
| 11 | Qwen 3 VL 8B (Thinking) | 16 |
| 12 | Qwen 2.5 VL 72B Instruct | 14.3 |
| 13 | GPT-5 | 14.2 |
| 14 | Qwen 3 VL 8B Instruct | 13.2 |
| 15 | InternVL3.5-8B | 11.3 |
Interactive version: theaggregate.ai/benchmark?slug=chartarena-combination-chart-english · How It Works · Data refreshed daily, snapshot 2026-09-29.