benchr
Leaderboard · Reasoning & Math
AI Reasoning Models Leaderboard
Ranked comparison of Large Language Models on reasoning, math, and problem-solving. Sourced from official docs, sorted by reasoning score.
| Rank | Model | Provider | Reasoning Rating (0–10) | GPQA Diamond | MATH |
|---|---|---|---|---|---|
| Loading leaderboard rankings... | |||||
Methodology
This leaderboard ranks models based on their **Reasoning capability rating (0–10)** from the benchr index, which synthesizes human-evaluation and math/logic performance. We also display **GPQA Diamond** (graduate-level science reasoning) and **MATH** scores for reference.