lupantech/MathVista

MathVista: data, code, and evaluation for Mathematical Reasoning in Visual Contexts

47
/ 100
Emerging

This project offers a comprehensive benchmark to evaluate how well AI models can solve math problems presented with visual information like charts, graphs, or diagrams. It takes a mathematical question combined with an image and assesses the model's ability to provide the correct numerical or logical answer. Scientists, researchers, and engineers working on advanced AI systems would use this to rigorously test and compare the mathematical reasoning capabilities of different AI models.

355 stars. No commits in the last 6 months.

Use this if you need to objectively measure and compare the performance of different large multimodal AI models on complex mathematical reasoning tasks involving visual data.

Not ideal if you are looking for an AI tool to solve your personal math problems or if you don't work with developing or evaluating AI models.

AI evaluation mathematical reasoning multimodal AI model benchmarking visual question answering
Stale 6m No Package No Dependents
Maintenance 2 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 19 / 25

How are scores calculated?

Stars

355

Forks

50

Language

Jupyter Notebook

License

CC-BY-SA-4.0

Last pushed

Sep 29, 2025

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/llm-tools/lupantech/MathVista"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.