Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | Claude 3.5 Haiku | Claude 3 Sonnet |
|---|---|---|
| GPQA Diamond | 41.6 | 40.4 |
coding
| Benchmark | Claude 3.5 Haiku | Claude 3 Sonnet |
|---|---|---|
| SWE-bench Verified | 40.6 | — |
agent
| Benchmark | Claude 3.5 Haiku | Claude 3 Sonnet |
|---|---|---|
| TAU-Bench Retail | 51 | — |
Category Winners
coding
6.5
reasoning
9.6
agent
10.1