Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | o1-preview | Sarvam-105B |
|---|---|---|
| GPQA Diamond | 73.3 | 78.7 |
| Humanity's Last Exam | — | 11.2 |
coding
| Benchmark | o1-preview | Sarvam-105B |
|---|---|---|
| SWE-bench Verified | 41.3 | 45 |
general
| Benchmark | o1-preview | Sarvam-105B |
|---|---|---|
| SimpleQA | 42.4 | — |
math
| Benchmark | o1-preview | Sarvam-105B |
|---|---|---|
| AIME 2025 | — | 96.7 |
agent
| Benchmark | o1-preview | Sarvam-105B |
|---|---|---|
| BrowseComp | — | 49.5 |
Category Winners
coding
S
Sarvam-105B10.0
math
S
Sarvam-105B80.2
reasoning
60.3
general
56.8
agent
S
Sarvam-105B21.2