Alibaba's powerful open-source model with strong multilingual performance, particularly for Chinese and other Asian languages.
| Benchmark | Description | Score | Category | Bar |
|---|---|---|---|---|
| MMLU | Massive Multitask Language Understanding | 84% | knowledge | |
| HumanEval | Code generation accuracy | 85.2% | coding | |
| MATH | Mathematical reasoning | 83.4% | math | |
| GSM8K | Grade school math | 86.1% | math | |
| GPQA | Graduate-level science Q&A | 44.5% | reasoning | |
| HellaSwag | Common sense reasoning | 84.2% | reasoning | |
| SWE-bench | Software engineering problems | 31.2% | coding | |
| ARC-C | AI2 Reasoning Challenge | 92% | reasoning |