Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | Kimi K2-Instruct-0905 | Magistral Medium |
|---|---|---|
| GPQA Diamond | 75.1 | 70.8 |
| Humanity's Last Exam | 4.7 | 9 |
math
| Benchmark | Kimi K2-Instruct-0905 | Magistral Medium |
|---|---|---|
| AIME 2025 | 49.5 | 64.9 |
coding
| Benchmark | Kimi K2-Instruct-0905 | Magistral Medium |
|---|---|---|
| SWE-bench Verified | 65.8 | — |
| TerminalBench | 25 | — |
general
| Benchmark | Kimi K2-Instruct-0905 | Magistral Medium |
|---|---|---|
| SimpleQA | 31 | — |
Category Winners
coding
M
Kimi K2-Instruct-090533.7
math
24.3
reasoning
M
Kimi K2-Instruct-090535.1
general
M
Kimi K2-Instruct-090539.3