Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | DeepSeek-V4-Flash-0423 | Gemini 3.5 Flash |
|---|---|---|
| GPQA Diamond | 87.4 | — |
| Humanity's Last Exam | 40.3 | 40.2 |
| ARC-AGI v2 | — | 72.1 |
coding
| Benchmark | DeepSeek-V4-Flash-0423 | Gemini 3.5 Flash |
|---|---|---|
| SWE-bench Verified | 78.6 | — |
| SWE-Bench Pro | 52.3 | 55.1 |
agent
| Benchmark | DeepSeek-V4-Flash-0423 | Gemini 3.5 Flash |
|---|---|---|
| BrowseComp | 53.5 | — |
| Toolathlon | 43.5 | 56.5 |
| MCP Atlas | 67.4 | 83.6 |
multimodal
| Benchmark | DeepSeek-V4-Flash-0423 | Gemini 3.5 Flash |
|---|---|---|
| MMMU-Pro | — | 83.6 |
| CharXiv Reasoning | — | 84.2 |
Category Winners
coding
54.4
reasoning
75.8
general
36.2
multimodal
80.4
agent
72.3