Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | DeepSeek-V4-Flash-0423 | Laguna S 2.1 |
|---|---|---|
| GPQA Diamond | 87.4 | — |
| Humanity's Last Exam | 40.3 | — |
coding
| Benchmark | DeepSeek-V4-Flash-0423 | Laguna S 2.1 |
|---|---|---|
| SWE-bench Verified | 78.6 | — |
| SWE-Bench Pro | 52.3 | 59.4 |
general
| Benchmark | DeepSeek-V4-Flash-0423 | Laguna S 2.1 |
|---|---|---|
| SimpleQA | 28.9 | — |
agent
| Benchmark | DeepSeek-V4-Flash-0423 | Laguna S 2.1 |
|---|---|---|
| BrowseComp | 53.5 | — |
| Toolathlon | 43.5 | 49.7 |
| MCP Atlas | 67.4 | — |
Category Winners
coding
P
Laguna S 2.160.4
reasoning
75.8
general
36.2
agent
P
Laguna S 2.139.0