Neura Intelligence Index
| Benchmark | Grok-3 | Seed 2.1 Turbo |
|---|---|---|
| GPQA Diamond | 84.6 | — |
| Humanity's Last Exam | — | 54.6 |
| Benchmark | Grok-3 | Seed 2.1 Turbo |
|---|---|---|
| AIME 2025 | 93.3 | — |
| Benchmark | Grok-3 | Seed 2.1 Turbo |
|---|---|---|
| MMMU | 78 | — |
| MMMU-Pro | — | 82.2 |
| CharXiv Reasoning | — | 83.6 |
| Benchmark | Grok-3 | Seed 2.1 Turbo |
|---|---|---|
| SWE-Bench Pro | — | 57 |
| SciCode | — | 57.8 |
| Benchmark | Grok-3 | Seed 2.1 Turbo |
|---|---|---|
| OSWorld | — | 76.4 |
| BrowseComp | — | 84.9 |
| Toolathlon | — | 49.1 |
| MCP Atlas | — | 80.3 |
| APEX Agents |
| — |
29.2 |