InfoVQA
multimodal
+
+
+
+
About
InfoVQA (Infographic Visual Question Answering) is a comprehensive benchmark featuring diverse infographics with natural language questions and answers. This dataset tests AI models' ability to understand complex visual information, extract text from images, and perform reasoning over infographic content. InfoVQA evaluates multimodal understanding capabilities for real-world information graphics and data visualizations.
+
+
+
+
Evaluation Stats
Total Models9
Organizations4
Verified Results0
Self-Reported9
+
+
+
+
Benchmark Details
Max Score1
Language
en
+
+
+
+
Performance Overview
Score distribution and top performers
Score Distribution
9 models
Top Score
83.4%
Average Score
71.6%
High Performers (80%+)
2Top Organizations
#1Alibaba Cloud / Qwen Team
2 models
83.0%
#2DeepSeek
3 models
73.3%
#3Microsoft
1 model
72.7%
#4Google
3 models
61.8%
+
+
+
+
Leaderboard
9 models ranked by performance on InfoVQA
| License | Links | ||||
|---|---|---|---|---|---|
| Feb 28, 2025 | Apache 2.0 | 83.4% | |||
| Jan 26, 2025 | Apache 2.0 | 82.6% | |||
| Dec 13, 2024 | deepseek | 78.1% | |||
| Dec 13, 2024 | deepseek | 75.8% | |||
| Feb 1, 2025 | MIT | 72.7% | |||
| Mar 12, 2025 | Gemma | 70.6% | |||
| Dec 13, 2024 | deepseek | 66.1% | |||
| Mar 12, 2025 | Gemma | 64.9% | |||
| Mar 12, 2025 | Gemma | 50.0% |
