Grok 3 Think (source unspecified)
xAI
This generative benchmark entry preserves results explicitly identified as Grok 3 (Think) or Grok 3 Beta Extended Thinking. xAI documents a separately trained consumer reasoning model, but the benchmark sources do not identify an immutable checkpoint or canonical API identifier. These observations are separate from the non-reasoning grok-3 API artifact.
Benchmark scores
| Benchmark | Score |
|---|---|
| AIME 2025 | 77.3 |
| GPQA Diamond | 80.2 |
| LiveCodeBench - October 2024 to February 2025 | 70.6 |
| MMMU | 76.0 |
| MRCR v2 (8-needle) - 128K average | 34.0 |
| USAMO 2025 | 4.8 |