Speechmatics - API for Voice AI Agents
Mini toolsMathVista
MathVista (testmini)
Mathematical reasoning over visual contexts: figures, charts, diagrams, geometric drawings.
Multimodal
Multimodal
accuracy
Max 100.0%
Released Oct 2023
8
Results
8
Models scored
86.8%
Top: o3
71.9%
Median
Best results
Top primary scores; one row per model.
Frontier over time
Each dot is one model result; the line traces the running best score.
All results
Showing all configurations including non-primary alternates. · Show only primary
| # | Model | Score | Conditions | Eval date | Source | Flags |
|---|---|---|---|---|---|---|
| 1 | o3 | 86.8% | — | Apr 16, 2025 | self reported | primary |
| 2 | o4 mini | 84.3% | — | Apr 16, 2025 | self reported | primary |
| 3 | Llama 4 Maverick | 73.7% | — | Apr 5, 2025 | self reported | primary |
| 4 | GPT 4.1 | 72.0% | — | Apr 14, 2025 | self reported | primary |
| 5 | o1 | 71.8% | — | Apr 16, 2025 | self reported | primary |
| 6 | Llama 4 Scout | 70.7% | — | Apr 5, 2025 | self reported | primary |
| 7 | Claude Sonnet 3.5 | 67.7% | 0-shot · standard | Jun 20, 2024 | self reported | |
| 8 | GPT-4o | 63.8% | — | Apr 16, 2025 | self reported | primary |
| 9 | Pixtral 12B | 58.3% | CoT | Oct 10, 2024 | self reported | primary |
| 10 | Claude Haiku 3 | 46.4% | 0-shot · standard | Mar 4, 2024 | self reported |
