MMLU-redux-2.0 leaderboard
As of November 14, 2026, among the results listed here, Kimi-K2-Base has the highest reported MMLU-redux-2.0 score, 90.2% (Base model evaluation results; EM), published by Moonshot AI. 1 model has a MMLU-redux-2.0 result in our records.
Scores as the model makers published them, not re-run by us; each row links to its announcement.
Best reported score
90.2%
Reported by Moonshot AI, July 11, 2025 · Base model evaluation results; EM
1 model ranked · 1 maker
| Rank | Model | Score | Setting | Reported by | Date |
|---|---|---|---|---|---|
| 1 | Kimi-K2-Base | 90.2% | Base model evaluation results; EM | Moonshot AI | July 11, 2025 (model release) |
How much to trust these numbers
- Source
- Reported by the model makers themselves. Not re-run by LLMBenchmark or anyone independent.
- Settings
- All ranked results use the same reported setting.
- Freshness
- Published July 11, 2025.
What MMLU-redux-2.0 measures
It tests models on 5,700 manually re-annotated questions across 57 subjects; the search results do not specify what its score means. Source: edinburgh-dawg/mmlu-redux-2.0 · Datasets at Hugging Face