Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | GLM-5.1 | Llama 3.1 405B |
|---|---|---|
| GPQA Diamond | 86.2 | 51.1 |
| Humanity's Last Exam | 52.3 | — |
| BigBench-Hard | — | 85.4 |
| ARC-Challenge | — | 96.9 |
coding
| Benchmark | GLM-5.1 | Llama 3.1 405B |
|---|---|---|
| SWE-Bench Pro | 58.4 | — |
| HumanEval | — | 89 |
agent
| Benchmark | GLM-5.1 | Llama 3.1 405B |
|---|---|---|
| BrowseComp | 79.3 | — |
| Toolathlon | 40.7 | — |
| MCP Atlas | 71.8 | — |
general
| Benchmark | GLM-5.1 | Llama 3.1 405B |
|---|---|---|
| MMLU-Pro | — | 73.3 |
| Chatbot Arena Elo | — | 1244 |
| IFEval | — | 86.4 |
language
| Benchmark | GLM-5.1 | Llama 3.1 405B |
|---|---|---|
| WinoGrande | — | 87.6 |
| HellaSwag | — | 96 |
safety
| Benchmark | GLM-5.1 | Llama 3.1 405B |
|---|---|---|
| TruthfulQA | — | 65.3 |
Category Winners
coding
69.3
math
72.0
reasoning
Z
GLM-5.184.7
general
67.9
language
87.7
safety
59.8
agent
Z
GLM-5.149.3