Loading page…
Preferred languages
More languages supported by Google Translate
Open the selector to load the current language catalog.
Translations are provided by Google Translate.
Loading page…
Preferred languages
More languages supported by Google Translate
Open the selector to load the current language catalog.
Translations are provided by Google Translate.
Mathematics benchmark
FrontierMath legacy aggregate. Every model the catalog carries a published FrontierMath (legacy) value for, ranked by that value.
Legacy FrontierMath values retained for historical model pages. This field is not used in current rankings because it can mix prior benchmark versions and slices. 8 models in the ModelScale catalog have a published value.
Each point is a model with a published FrontierMath (legacy) value and complete token prices, on the 75% input / 25% output blend. Frontier models are spaced evenly along the cost axis.
8 models with a published FrontierMath (legacy) value, ordered by that value, highest first. Models the source has not scored on this benchmark are not listed — they are unmeasured, not last.
| Rank | Model | Provider | Open-ended mathematical reasoning with tool access |
|---|---|---|---|
| 1 | OpenAI | 89 | |
| 2 | OpenAI | 84.9 | |
| 3 | OpenAI | 78.6 |
What the benchmark measures and how its value is published, from the source’s own record.
FrontierMath (legacy) (FrontierMath legacy aggregate) is a mathematics benchmark published by FrontierMath: A Benchmark for Evaluating Advanced Mathematical Reasoning in AI. Legacy FrontierMath values retained for historical model pages. This field is not used in current rankings because it can mix prior benchmark versions and slices. Its evaluation set is described by the source as historical aggregate. The source rates its difficulty as research-level mathematics.
Each model's result is published by FrontierMath: A Benchmark for Evaluating Advanced Mathematical Reasoning in AI as open-ended mathematical reasoning with tool access. ModelScale prints that value exactly as published, and never converts or rescales it. This page ranks the 8 models that carry a published value, highest value first; a model the source never scored is left out rather than ranked at zero. The source does not state whether a higher value is the better result, so neither does this page.
| 4 | OpenAI | 52.4 |
| 5 | OpenAI | 51.7 |
| 6 | OpenAI | 50 |
| 7 | Anthropic | 43.8 |
| 7 | Anthropic | 43.8 |
Evidence key: ObservedLast good
Rows are ordered by the value FrontierMath: A Benchmark for Evaluating Advanced Mathematical Reasoning in AI published, highest first. The source does not state whether a higher value is the better result, so this page does not either: for a benchmark that measures a rate of failure, read the table from the bottom.