# MathVista — AI model rankings

mathematical reasoning about visual information

12 tracked models have a published MathVista score. Higher is better. Scores come from published lab reports and benchmark sources; recorded sources appear beside the scores.

## Ranking

| Rank | Model | Developer | Score | Source | Released |
| --- | --- | --- | --- | --- | --- |
| 1 | LLaMA 4 Maverick | Meta | 73.7% | Lab | Apr 5 2025 |
| 2 | GPT-4.1 mini | OpenAI | 73.1% | Lab | Apr 14 2025 |
| 3 | GPT-4.1 | OpenAI | 72.2% | Lab | Apr 14 2025 |
| 4 | LLaMA 4 Scout | Meta | 70.7% | Lab | Apr 5 2025 |
| 5 | Gemma 3 | Google | 67.6% | Lab | Mar 12 2025 |
| 6 | Gemini 1.5 Flash | Google | 58.4% | Lab | May 14 2024 |
| 7 | LLaMA 3.2 | Meta | 57.3% | Lab | Sep 25 2024 |
| 8 | GPT-4.1 nano | OpenAI | 56.2% | Lab | Apr 14 2025 |
| 9 | Gemini 1.0 Ultra | Google | 53% | Lab | Feb 8 2024 |
| 10 | Gemini 1.5 Pro | Google | 52.1% | Lab | Feb 15 2024 |
| 11 | Gemini 1.0 Pro | Google | 45.2% | Lab | Dec 13 2023 |
| 12 | Gemini 1.0 Nano | Google | 30.6% | Lab | Dec 6 2023 |


---

Canonical page: https://aireleasetracker.com/benchmark/mathvista
Site index: https://aireleasetracker.com/llms.txt
Source: AI Release Tracker (https://aireleasetracker.com). Most benchmark scores come from lab launch material; gathered results identify the leaderboard that published them.