Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | DeepSeek-V4.1-Flash-1017 | DeepSeek-V4-Pro-Max |
|---|---|---|
| Humanity's Last Exam | 42.6 | 48.2 |
| GPQA Diamond | — | 90.1 |
coding
| Benchmark | DeepSeek-V4.1-Flash-1017 | DeepSeek-V4-Pro-Max |
|---|---|---|
| SWE-bench Verified | — | 80.6 |
| SWE-Bench Pro | — | 55.4 |
general
| Benchmark | DeepSeek-V4.1-Flash-1017 | DeepSeek-V4-Pro-Max |
|---|---|---|
| SimpleQA | — | 57.9 |
agent
| Benchmark | DeepSeek-V4.1-Flash-1017 | DeepSeek-V4-Pro-Max |
|---|---|---|
| BrowseComp | — | 83.4 |
| Toolathlon | — | 51.8 |
| MCP Atlas | — | 73.6 |
Category Winners
coding
60.7
reasoning
84.3
general
78.1
agent
62.7