New: connect Claude & other AIs to GenAIList over MCP — research the catalog and contribute to the shared knowledge base. Learn how →
ChartQA benchmark
AI model leaderboard for the ChartQA benchmark. Compare how large language models score on ChartQA, see the full ranking, and understand what this AI benchmark measures. Claude 3.5 Sonnet currently leads with 90.8%. Visual question-answering over real-world charts and infographics.
Leaderboard
| # | Model | Organization | Score | Variant | Source |
|---|---|---|---|---|---|
| #1 | Claude 3.5 Sonnet | Anthropic | 90.8% | cited-nvlm | official ↗ |
| #2 | Claude 3.5 Sonnet | Anthropic | 89.1% | cot-cited-pixtral-large | official ↗ |
| #3 | Pixtral Large | Mistral AI | 88.1% | cot-from-hf-readme | official ↗ |
| #4 | Mistral Small 3.2 | Mistral AI | 87.4% | instruct-from-hf-readme | official ↗ |
| #5 | Gemini 1.5 Pro | Google DeepMind | 87.2% | aug-2024-cited-nvlm | official ↗ |
| #6 | Aria | Rhymes AI | 86.4% | from-hf-readme | official ↗ |
| #7 | Mistral Small 3.1 | Mistral AI | 86.24% | instruct-cited-msmall32 | official ↗ |
| #8 | NVLM-D 72B | NVIDIA | 86% | megatron-from-hf-readme | official ↗ |
| #9 | GPT-4o (Aug 2024) | OpenAI | 85.7% | cited-nvlm | official ↗ |
| #10 | Llama 3.2 90B | Meta AI | 85.5% | cited-nvlm | official ↗ |
| #11 | GPT-4o (Aug 2024) | OpenAI | 85.2% | cot-cited-pixtral-large | official ↗ |
| #12 | Gemini 1.5 Pro | Google DeepMind | 83.8% | cot-cited-pixtral-large | official ↗ |
| #13 | Llama 3.2 11B | Meta AI | 83.4% | cited-aria | official ↗ |
| #14 | Pixtral 12B | Mistral AI | 81.8% | cited-aria | official ↗ |
| #15 | GPT-4V | OpenAI | 78.5% | cited-nvlm | official ↗ |
| #16 | Llama 3.2 90B | Meta AI | 70.8% | cited-pixtral-large | official ↗ |
Frequently asked questions about ChartQA
What is the ChartQA benchmark?
Visual question-answering over real-world charts and infographics.
How is the ChartQA benchmark scored?
ChartQA is scored using the accuracy (%) metric, where a higher score is better. The maximum achievable score is 100.000. GenAIList aggregates reported scores from model providers and papers into a single ranked leaderboard.
Which AI model scores highest on ChartQA?
As of the latest reported scores on GenAIList, Claude 3.5 Sonnet achieves the highest result on ChartQA with a score of 90.8%.
Is a higher ChartQA score better?
Yes. On ChartQA a higher score indicates better performance, so models near the top of the leaderboard are the strongest.