Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | Claude Sonnet 4 | Qwen3-235B-A22B-Instruct-2507 |
|---|---|---|
| GPQA Diamond | 75.4 | 77.5 |
math
| Benchmark | Claude Sonnet 4 | Qwen3-235B-A22B-Instruct-2507 |
|---|---|---|
| AIME 2025 | 70.5 | 70.3 |
coding
| Benchmark | Claude Sonnet 4 | Qwen3-235B-A22B-Instruct-2507 |
|---|---|---|
| SWE-bench Verified | 72.7 | — |
| TerminalBench | 35.5 | — |
multimodal
| Benchmark | Claude Sonnet 4 | Qwen3-235B-A22B-Instruct-2507 |
|---|---|---|
| MMMU | 74.4 | — |
agent
| Benchmark | Claude Sonnet 4 | Qwen3-235B-A22B-Instruct-2507 |
|---|---|---|
| TAU-Bench Retail | 80.5 | — |
Category Winners
coding
60.0
math
34.1
reasoning
Q
Qwen3-235B-A22B-Instruct-250770.6
general
Q
Qwen3-235B-A22B-Instruct-250773.3
multimodal
68.6
agent
83.4