DataLearner 标志

DeepSeek-V4.1-FlashvsDeepSeek-V4-Flash-Vision-Exp

在 8 个共同 benchmark 中,DeepSeek-V4.1-Flash 整体领先:DeepSeek-V4.1-Flash 领先 8 项,DeepSeek-V4-Flash-Vision-Exp 领先 0 项,持平 0 项,平均分差 +13.04。

DeepSeek-AI
DeepSeek-V4.1-Flash

DeepSeek-AI · 2026-09-10 · 多模态大模型

DeepSeek-AI
DeepSeek-V4-Flash-Vision-Exp

DeepSeek-AI · 2026-08-21 · 多模态大模型

DeepSeek-V4.1-Flash8 (100%)(0%)0 DeepSeek-V4-Flash-Vision-Exp

评测分数

按能力类目分组,每组内按分差大小排列;共 8 项。

AI Agent - Tool Usage

DeepSeek-V4.1-Flash 领先 2/2
评测项DeepSeek-V4.1-FlashDeepSeek-V4-Flash-Vision-Exp分差
CyberGym88.101 / 8Max (With Tools)75.308 / 8Max (With Tools)+12.80
Terminal-Bench 2.190.601 / 53Max (With Tools)83.9018 / 53Max (With Tools)+6.70

Coding and Software Engineer

DeepSeek-V4.1-Flash 领先 2/2
评测项DeepSeek-V4.1-FlashDeepSeek-V4-Flash-Vision-Exp分差
DeepSWE74.202 / 38Max (With Tools)59.3021 / 38Max (With Tools)+14.90
NL2Repo-Bench65.402 / 16Max (With Tools)57.708 / 16Max (With Tools)+7.70

Multimodal Understanding

DeepSeek-V4.1-Flash 领先 2/2
评测项DeepSeek-V4.1-FlashDeepSeek-V4-Flash-Vision-Exp分差
Chartography78.904 / 8Max (With Tools)64.307 / 8Max (With Tools)+14.60
ZeroBench Main493 / 6Max (With Tools)355 / 6Max (With Tools)+14

Agent Level Benchmark

DeepSeek-V4.1-Flash 领先 1/1
评测项DeepSeek-V4.1-FlashDeepSeek-V4-Flash-Vision-Exp分差
Agents' Last Exam31.806 / 19Max (With Tools)27.3010 / 19Max (With Tools)+4.50

Productivity Knowledge

DeepSeek-V4.1-Flash 领先 1/1
评测项DeepSeek-V4.1-FlashDeepSeek-V4-Flash-Vision-Exp分差
AutomationBench54.801 / 17Max (With Tools)25.7016 / 17Max (With Tools)+29.10

规格对比

字段DeepSeek-V4.1-FlashDeepSeek-V4-Flash-Vision-Exp
发布机构DeepSeek-AIDeepSeek-AI
发布时间2026-09-102026-08-21
模型类型多模态大模型多模态大模型
架构MoE 架构MoE 架构
参数规模5520亿3050亿
上下文长度1M1M
最大输出384K384K

API 调用价格

价格优先使用 DataLearner 配置的 API 记录;缺失项不做推测。

价格项DeepSeek-V4.1-FlashDeepSeek-V4-Flash-Vision-Exp
文本输入¥1 / 1M tokens$0.22 / 1M tokens
文本输出¥4 / 1M tokens$0.66 / 1M tokens
缓存读取¥0.02 / 1M tokens$0.007 / 1M tokens

小结

  • DeepSeek-V4.1-Flash在以下类目领先:AI Agent - Tool Usage (2/2)、Coding and Software Engineer (2/2)、Multimodal Understanding (2/2)、Agent Level Benchmark (1/1)、Productivity Knowledge (1/1)

8 个共同 benchmark 上,DeepSeek-V4.1-Flash 平均高出 13.04 分。

单项差距最大的 benchmark:AutomationBench — DeepSeek-V4.1-Flash 54.80,DeepSeek-V4-Flash-Vision-Exp 25.70(分差 +29.10)。

本页正文由结构化模型、价格与 benchmark 数据生成,不使用实时 LLM 撰写。