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.
Coding benchmark
Every model the catalog carries a published LiveCodeBench Pro value for, ranked by that value.
A harder competitive-programming benchmark family built from Codeforces, ICPC, and IOI problems, with quarter-specific public leaderboards and difficulty-aware reporting. 9 models in the ModelScale catalog have a published value.
Each point is a model with a published LiveCodeBench Pro value and complete token prices, on the 75% input / 25% output blend. Frontier models are spaced evenly along the cost axis.
9 models with a published LiveCodeBench Pro 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 | Competitive programming |
|---|---|---|---|
| 1 | SASakana Fugu-Ultra | Sakana AI | 90.8 |
| 2 | SASakana Fugu | Sakana AI | 87.8 |
| 3 | OpenAI | 87.5 | |
| 4 |
What the benchmark measures and how its value is published, from the source’s own record.
LiveCodeBench Pro is a coding benchmark published by LiveCodeBench Pro: How Do Olympiad Medalists Judge LLMs in Competitive Programming. A harder competitive-programming benchmark family built from Codeforces, ICPC, and IOI problems, with quarter-specific public leaderboards and difficulty-aware reporting. Its evaluation set is described by the source as quarter-specific contest programming sets. The source rates its difficulty as high-end contest programming.
Each model's result is published by LiveCodeBench Pro: How Do Olympiad Medalists Judge LLMs in Competitive Programming as competitive programming. ModelScale prints that value to 1 decimal place, exactly as published, and never converts or rescales it. This page ranks the 9 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.
| 82.9 |
| 5 | Meta | 80.0 |
| 6 | xAI | 74.2 |
| 7 | Anthropic | 70.7 |
| 8 | LiquidAI | 31.7 |
| 9 | OPMiniCPM5-1B | OpenBMB | 22.7 |
Evidence key: Observed
Rows are ordered by the value LiveCodeBench Pro: How Do Olympiad Medalists Judge LLMs in Competitive Programming? 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.