Llama 3.3 is a multilingual large language model optimized for dialogue use cases across multiple languages. It is a pretrained and instruction-tuned generative model with 70 billion parameters, outperforming many open-source and closed chat models on common industry benchmarks. Llama 3.3 supports a context length of 128,000 tokens and is designed for commercial and research use in multiple languages.
发布日期2024年12月6日
参数规模70B
上下文长度131K
许可证Llama 3.3 Community License Agreement
知识截止—
Benchmarks
评测成绩
| 评测基准 | 类别 | 分数 | 来源 |
|---|---|---|---|
| IFEval | general | 92.1 | 来源 |
| MGSM | math reasoning | 91.1 | 来源 |
| HumanEval | reasoning code | 88.4 | 来源 |
| MBPP EvalPlus | reasoning general | 87.6 | 来源 |
| MMLU | general reasoning language math | 86.0 | 来源 |
| BFCL v2 | general reasoning | 77.3 | 来源 |
| MATH | math reasoning | 77.0 | 来源 |
| MMLU-Pro | language reasoning math general | 68.9 | 来源 |
| GPQA | reasoning general | 50.5 | 来源 |
Pricing
API 价格对比
| 服务商 | 输入价 | 输出价 | 上下文 | 吞吐(tok/s) | 延迟(s) | 函数调用 | 代码执行 | 联网搜索 |
|---|---|---|---|---|---|---|---|---|
| Llama | $0.00 | $0.00 | 128K | — | — | ✓ | ✗ | ✗ |
| Cortecs | $0.09 | $0.28 | 131K | — | — | ✓ | ✗ | ✗ |
| Helicone | $0.13 | $0.39 | 128K | — | — | ✓ | ✗ | ✗ |
| LLM Gateway | $0.13 | $0.40 | 131K | — | — | ✓ | ✗ | ✗ |
| Regolo AI | $0.60 | $2.70 | 128K | — | — | ✓ | ✗ | ✗ |
| Azure | $0.71 | $0.71 | 128K | — | — | ✓ | ✗ | ✗ |
| Azure Cognitive Services | $0.71 | $0.71 | 128K | — | — | ✓ | ✗ | ✗ |
| Scaleway | $0.90 | $0.90 | 100K | — | — | ✓ | ✗ | ✗ |
价格单位：美元/百万 token，数据来自社区整理，仅供参考。