Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | DeepSeek R1 Distill Llama 70B | Qwen3-Next-80B-A3B-Instruct |
|---|---|---|
| GPQA Diamond | 65.2 | 72.9 |
math
| Benchmark | DeepSeek R1 Distill Llama 70B | Qwen3-Next-80B-A3B-Instruct |
|---|---|---|
| AIME 2025 | — | 69.5 |
agent
| Benchmark | DeepSeek R1 Distill Llama 70B | Qwen3-Next-80B-A3B-Instruct |
|---|---|---|
| TAU-Bench Retail | — | 60.9 |
Category Winners
math
Q
Qwen3-Next-80B-A3B-Instruct32.4
reasoning
Q
Qwen3-Next-80B-A3B-Instruct62.2
agent
Q
Qwen3-Next-80B-A3B-Instruct30.0