Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | DeepSeek-R1-0528 | QwQ-32B |
|---|---|---|
| GPQA Diamond | 81 | 65.2 |
| Humanity's Last Exam | 17.7 | — |
math
| Benchmark | DeepSeek-R1-0528 | QwQ-32B |
|---|---|---|
| AIME 2025 | 87.5 | — |
coding
| Benchmark | DeepSeek-R1-0528 | QwQ-32B |
|---|---|---|
| SWE-bench Verified | 44.6 | — |
| TerminalBench | 5.7 | — |
general
| Benchmark | DeepSeek-R1-0528 | QwQ-32B |
|---|---|---|
| SimpleQA | 92.3 | — |
agent
| Benchmark | DeepSeek-R1-0528 | QwQ-32B |
|---|---|---|
| BrowseComp | 8.9 | — |
Category Winners
coding
5.5
math
65.5
reasoning
50.0
general
98.2
agent
0.3