Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
agent
| Benchmark | OpenAI Codex | Seed 1.8 |
|---|---|---|
| GAIA | 68.5 | — |
| WebArena | 32 | — |
| OSWorld | — | 61.9 |
| BrowseComp | — | 67.6 |
coding
| Benchmark | OpenAI Codex | Seed 1.8 |
|---|---|---|
| SWE-bench Verified | 69.1 | 72.9 |
math
| Benchmark | OpenAI Codex | Seed 1.8 |
|---|---|---|
| AIME 2025 | — | 94.3 |
multimodal
| Benchmark | OpenAI Codex | Seed 1.8 |
|---|---|---|
| MMMU | — | 83.4 |
| MMMU-Pro | — | 73.2 |
| CharXiv Reasoning | — | 71.4 |
Category Winners
coding
B
Seed 1.867.2
math
B
Seed 1.876.8
multimodal
B
Seed 1.863.7
agent
78.2