DeepSeek VL2
DeepSeekDeepSeek开源权重deepseek
描述
An advanced series of large Mixture-of-Experts (MoE) Vision-Language Models that significantly improves upon its predecessor, DeepSeek-VL. DeepSeek-VL2 demonstrates superior capabilities across various tasks, including but not limited to visual question answering, optical character recognition, document/table/chart understanding, and visual grounding.
发布日期
2024-12-13
参数规模
27.0B
上下文长度
—
支持模态
image, text
能力雷达图
70
general
0
coding
60
reasoning
43
science估算
42
agents
90
multimodal
缺少专门科学评测时，Science 由 LLM Stats 科学得分或推理能力估算。
排行榜排名
| 领域 | #排名 | 分数 | 来源 |
|---|---|---|---|
| 多模态榜 | 79 | 28.0 | LS |
基准测试分数 (LLM Stats)
(LLM Stats (zeroeval))General
MMT-Bench
63.6%自报
MMStar
61.3%自报
MMMU
51.1%自报
Image To Text
DocVQADocVQA (2020)
93.3%自报
TextVQA
84.2%自报
OCRBench
81.1%自报
Math
MathVista
62.8%自报
Multimodal
ChartQAMasry et al. (2022)
86.0%自报
AI2D
81.4%自报
MMBench
79.6%自报
MMBench-V1.1
79.2%自报
InfoVQA
78.1%自报
MME
22.5%自报
Spatial Reasoning
RealWorldQA
68.4%自报
AA 评测指数
(Artificial Analysis)暂无 AA 评测数据
LLM Stats 分类评分
(LLM Stats (zeroeval))Image To Text90
Multimodal70
Reasoning70
Spatial Reasoning70
General70
Vision70
Math60
Healthcare50
定价
暂无定价数据
速度
暂无速度数据
供应商价格排行
暂无提供商数据