benchmark evidence
MathVista
MathVista: math reasoning with visual context — diagrams, charts, geometry, visual equations.
activeupstream source ->
winner on MathVista
OpenAI: o386.8
direct benchmark result, not a broad vertical composite | source row dated 2000-01-01
scored on 2000-01-01 · stale source data (9650d)
latest mapped results | top 20
| # | Model | Score | Evidence | Tested |
|---|---|---|---|---|
| 1 | OpenAI: o3 | 86.8 | model-only independent_benchmark | 2000-01-01 |
| 2 | OpenAI: o4 Mini | 84.3 | model-only independent_benchmark | 2000-01-01 |
| 3 | Meta: Llama 4 Maverick | 73.7 | model-only independent_benchmark | 2000-01-01 |
| 4 | OpenAI: GPT-4.1 | 72.2 | model-only independent_benchmark | 2000-01-01 |
| 5 | OpenAI: o1 | 71.8 | model-only independent_benchmark | 2000-01-01 |
| 6 | Meta: Llama 4 Scout | 70.7 | model-only independent_benchmark | 2000-01-01 |
what this result means
MathVista: math reasoning with visual context — diagrams, charts, geometry, visual equations.
This benchmark contributes direct public evidence. Read its scope before generalizing the result.
A win here is a win on MathVista. Broad task pages require independent corroboration before naming a general winner.
source record
category: multimodal
metric: accuracy
matched models: 6
latest source date: 2000-01-01
direction: higher is better