Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
coding
| Benchmark | DeepSeek-V2.5 | Gemma 3n E4B Instructed LiteRT Preview |
|---|---|---|
| SWE-bench Verified | 16.8 | — |
reasoning
| Benchmark | DeepSeek-V2.5 | Gemma 3n E4B Instructed LiteRT Preview |
|---|---|---|
| GPQA Diamond | — | 23.7 |
math
| Benchmark | DeepSeek-V2.5 | Gemma 3n E4B Instructed LiteRT Preview |
|---|---|---|
| AIME 2025 | — | 11.6 |
Category Winners
coding
0.1
math
0.1
reasoning
1.3