Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | DeepSeek R1 Distill Llama 70B | GPT OSS 120B |
|---|---|---|
| GPQA Diamond | 65.2 | 80.1 |
| Humanity's Last Exam | — | 14.9 |
agent
| Benchmark | DeepSeek R1 Distill Llama 70B | GPT OSS 120B |
|---|---|---|
| TAU-Bench Retail | — | 67.8 |
Category Winners
reasoning
47.0
agent
50.1