ChartQA Pro leaderboard
ChartQA Pro
3 models tested · Updated 2026-04-14 · Verified sources only
Qwen3.5 27B leads at 66.8%
1
Qwen3.5 27B
Qwen · arxiv/2604.08644 · 2026-04-14
From EXAONE 4.5 technical report comparison. Best among 27-33B models.
66.8%
2
EXAONE 4.5 33B
LG AI Research · HuggingFace/LGAI-EXAONE · 2026-04-14
Highest ChartQA Pro score in its class. 33B dense model beating GPT-5 mini (60.9) and Qwen3-VL 32B Thinking (61.4).
62.2%
3
GPT-5 mini
OpenAI · arxiv/2604.08644 · 2026-04-14
From EXAONE 4.5 technical report comparison. Trails EXAONE 4.5 33B (62.2) and Qwen3.5 27B (66.8).
60.9%
benchmark.space canonical public ledger of AI capability