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.
Multilingual benchmark
Multilingual Grade School Math. Every model the catalog carries a published MGSM value for, ranked by that value.
A multilingual benchmark that translates 250 grade school math problems from GSM8K into 10 typologically diverse languages: Bengali, German, Spanish, French, Japanese, Russian, Swahili, Telugu, Thai, and Chinese. 13 models in the ModelChoice catalog have a published value.
Each point is a model with a published MGSM value and complete token prices, on the 75% input / 25% output blend. Frontier models are spaced evenly along the cost axis.
13 models with a published MGSM 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 | Math word problems |
|---|---|---|---|
| 1 | OpenAI | 93.7 | |
| 2 | Meta | 91.1 | |
| 3 | OpenAI | 89.9 | |
| 4 |
What the benchmark measures and how its value is published, from the source’s own record.
MGSM (Multilingual Grade School Math) is a multilingual benchmark published by Language Models are Multilingual Chain-of-Thought Reasoners. A multilingual benchmark that translates 250 grade school math problems from GSM8K into 10 typologically diverse languages: Bengali, German, Spanish, French, Japanese, Russian, Swahili, Telugu, Thai, and Chinese. Its evaluation set is described by the source as 250 problems × 11 languages. The source rates its difficulty as grade school math, multilingual.
Each model's result is published by Language Models are Multilingual Chain-of-Thought Reasoners as math word problems. ModelChoice prints that value exactly as published, and never converts or rescales it. This page ranks the 13 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.
| OpenAI |
| 86.9 |
| 4 | Meta | 86.9 |
| 6 | Anthropic | 85.6 |
| 7 | Microsoft | 80.6 |
| 8 | Meta | 68.9 |
| 9 | Microsoft | 47.9 |
| 10 | IBM | 37.84 |
| 11 | IBM | 27.52 |
| 12 | IBM | 16.16 |
| 13 | IBM | 14.72 |
Evidence key: Observed
Rows are ordered by the value Language Models are Multilingual Chain-of-Thought Reasoners 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.