Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | DeepSeek-V4-Flash-Max | Gemini 3.6 Flash |
|---|---|---|
| GPQA Diamond | 88.1 | — |
| Humanity's Last Exam | 45.1 | — |
coding
| Benchmark | DeepSeek-V4-Flash-Max | Gemini 3.6 Flash |
|---|---|---|
| SWE-bench Verified | 79 | — |
| SWE-Bench Pro | 52.6 | 58.7 |
agent
| Benchmark | DeepSeek-V4-Flash-Max | Gemini 3.6 Flash |
|---|---|---|
| BrowseComp | 73.2 | — |
| Toolathlon | 47.8 | — |
| MCP Atlas | 69 | — |
multimodal
| Benchmark | DeepSeek-V4-Flash-Max | Gemini 3.6 Flash |
|---|---|---|
| CharXiv Reasoning | — | 89.4 |
Category Winners
coding
54.5
reasoning
80.9
general
56.1
multimodal
84.1
agent
48.4