MathVista: Results as Published

MathVista tests mathematical reasoning that depends on a picture, including figures, charts, geometry diagrams, function plots, tables, scientific figures, and natural images. The full set has 6,141 examples, drawn from 28 existing multimodal datasets plus three new ones, IQTest, FunctionQA, and PaperQA. Most reported results use testmini, a 1,000-example subset; the remaining 5,141 examples are the held-out test set. The published result is accuracy.

Every number here is read from the same catalog file as the benchmark table, data version 2026103001, and is the number the table shows for that model and test. Where one page printed a model under several setups, each value is listed. The catalog is public in armanobosyan/sugra-llm-benchmarks; a wrong number can be reported from the table, with Report an error under the value, or as an issue in that repository.