跳转到主要内容

DeepSeek VL2

DeepSeekDeepSeekOpen Weightdeepseek

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
164K
支持模态
image, text

能力雷达图

60
general
0
coding
60
reasoning
43
science估算
0
agents
90
multimodal

Science 在缺少专门科学评测时使用推理能力代理估算。

排行榜排名

领域#排名分数来源
多模态榜42
76.0
LS

基准测试分数 (LLM Stats)

General

MMT-Bench63.6%自报
MMStar61.3%自报
MMMU51.1%自报

Image To Text

DocVQA93.3%自报
TextVQA84.2%自报
OCRBench81.1%自报

Math

MathVista62.8%自报

Multimodal

ChartQA86.0%自报
AI2D81.4%自报
MMBench79.6%自报
MMBench-V1.179.2%自报
InfoVQA78.1%自报
MME22.5%自报

Spatial Reasoning

RealWorldQA68.4%自报

AA 评测指数

暂无 AA 评测数据

LLM Stats 分类评分

Image To Text
90
Spatial Reasoning
70
Vision
70
Multimodal
70
Reasoning
70
General
60
Math
60
Healthcare
50

定价

输入价格$0.32 / 1M tokens
输出价格$0.89 / 1M tokens
混合价格(3:1)$0.4625 / 1M tokens

速度

暂无速度数据

可用提供商

(LS 内部计价单位)

暂无提供商数据

外部链接