A low-cost multimodal model that is lightning fast for processing images, video, documents, and text.
发布日期2024年11月20日
参数规模—
上下文长度—
许可证Proprietary
知识截止—
Benchmarks
评测成绩
| 评测基准 | 类别 | 分数 | 来源 |
|---|---|---|---|
| GSM8k | math reasoning | 94.5 | 来源 |
| ARC-C | reasoning general | 92.4 | 来源 |
| DocVQA | vision multimodal | 92.4 | 来源 |
| IFEval | general | 89.7 | 来源 |
| Translation Set1→en COMET22 | language | 88.8 | 来源 |
| Translation en→Set1 COMET22 | language | 88.8 | 来源 |
| ChartQA | reasoning vision multimodal | 86.8 | 来源 |
| HumanEval | reasoning code | 85.4 | 来源 |
| BBH | reasoning math language | 82.4 | 来源 |
| MMLU | general reasoning language math | 80.5 | 来源 |
| GroundUI-1K | multimodal vision | 80.2 | 来源 |
| TextVQA | vision multimodal image-to-text | 80.2 | 来源 |
| DROP | reasoning math | 80.2 | 来源 |
| VATEX | multimodal video language | 77.8 | 来源 |
| VisualWebBench | vision multimodal frontend_development | 77.7 | 来源 |
| FinQA | finance math reasoning | 73.6 | 来源 |
| MATH | math reasoning | 73.3 | 来源 |
| EgoSchema | vision reasoning long_context | 71.4 | 来源 |
| BFCL | general reasoning | 66.6 | 来源 |
| MM-Mind2Web | multimodal frontend_development reasoning | 60.7 | 来源 |
| MMMU | multimodal reasoning general | 56.2 | 来源 |
| CRAG | reasoning search | 43.8 | 来源 |
| Translation Set1→en spBleu | language | 43.1 | 来源 |
| GPQA | reasoning general | 42.0 | 来源 |
| Translation en→Set1 spBleu | language | 41.5 | 来源 |
| LVBench | vision multimodal long_context | 40.4 | 来源 |
| SQuALITY | summarization long_context language | 19.2 | 来源 |
Pricing
API 价格对比
暂无 API 价格。