Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | GPT OSS 120B | LongCat-Flash-Chat |
|---|---|---|
| GPQA Diamond | 80.1 | 73.2 |
| Humanity's Last Exam | 14.9 | — |
agent
| Benchmark | GPT OSS 120B | LongCat-Flash-Chat |
|---|---|---|
| TAU-Bench Retail | 67.8 | — |
math
| Benchmark | GPT OSS 120B | LongCat-Flash-Chat |
|---|---|---|
| AIME 2025 | — | 61.3 |
coding
| Benchmark | GPT OSS 120B | LongCat-Flash-Chat |
|---|---|---|
| SWE-bench Verified | — | 60.4 |
| TerminalBench | — | 39.5 |
Category Winners
coding
M
LongCat-Flash-Chat52.2
math
M
LongCat-Flash-Chat19.1
reasoning
M
LongCat-Flash-Chat60.2
agent
50.1