所有发布••

Open weights release

•

Released 2026年9月

DeepSeek V4.1 Flash:本次发布的智能、性能与价格

模型摘要

智能

39
Artificial Analysis Intelligence Index
智能为 4 档中的第 4 档。

速度

179.2
每秒输出 token 数
速度为 4 档中的第 4 档。

成本

输入 US$0.30输出 US$1.20缓存折扣 98%
$0.27
Intelligence Index 每项任务的成本
成本为 4 档中的第 3 档。

冗长度

250M
Intelligence Index 输出 token 数
冗长度为 4 档中的第 4 档。

DeepSeek V4.1 Flash 发布包含 2 个模型,各自具有不同的智能、性能与价格特征。下方对这 2 个模型的关键指标进行了比较。

  • 智能方面,DeepSeek V4.1 Flash 中得分最高的模型是 DeepSeek V4.1 Flash (Max)(39)。
  • 输出速度方面,最快的模型是 DeepSeek V4.1 Flash (Max)(179 token/秒)。
  • 延迟方面,DeepSeek V4.1 Flash (Non-reasoning)(1.08 秒) 的首个答案 token 等待时间最短。
  • 价格方面,DeepSeek V4.1 Flash (Non-reasoning)($0.15) 的单任务成本最低。不同模型的价格差异最高可达 1.8 倍。

Artificial Analysis Intelligence Index · Higher is better
Output tokens per second · Higher is better
Weighted average cost (USD) per Intelligence Index task · Lower is better

智能

Artificial Analysis Intelligence Index

Artificial Analysis Intelligence Index v4.3.2 incorporates 10 evaluations: AA-Briefcase v1.1, GDPval-AA v2.1, AutomationBench-AA, Terminal-Bench 4.0, SciCode, Humanity's Last Exam, GDP.pdf, CritPt, AA-Omniscience, AA-LCR v1.1

Intelligence Index 与每项任务的成本

Artificial Analysis Intelligence Index · Weighted average cost (USD) per Artificial Analysis Intelligence Index task
Most attractive quadrant
DeepSeek V4.1 Flash
Pareto line

成本

每项 Intelligence Index 任务的成本

Weighted average cost (USD) per Intelligence Index task · Segmented by token type · Lower is better

速度与延迟

输出速度

Output tokens per second · Higher is better

能力得分

能力指数

衡量模型在特定能力与行业中的表现
Finance & Accounting Index

Incorporates 7 evaluations: AA-Omniscience, GDPval-AA v2.1, AA-Briefcase v1.1, Humanity's Last Exam, AutomationBench-AA, AA-LCR v1.1, GDP.pdf · Higher is better

Strategy & Ops Index

Incorporates 6 evaluations: GDPval-AA v2.1, AA-Briefcase v1.1, AA-Omniscience, AutomationBench-AA, AA-LCR v1.1, GDP.pdf · Higher is better

Legal Index

Incorporates 7 evaluations: AA-Omniscience, GDPval-AA v2.1, AA-Briefcase v1.1, Humanity's Last Exam, AA-LCR v1.1, GDP.pdf, AutomationBench-AA · Higher is better

Healthcare & Medical Index

Incorporates 6 evaluations: AA-Omniscience, GDPval-AA v2.1, AA-Briefcase v1.1, MLCR-AA, Humanity's Last Exam, AutomationBench-AA · Higher is better

Engineering Index

Incorporates 6 evaluations: AA-Omniscience, Humanity's Last Exam, CritPt, GDPval-AA v2.1, AA-Briefcase v1.1, Terminal-Bench 4.0 · Higher is better

Economics Index

Incorporates 5 evaluations: AA-Omniscience, Humanity's Last Exam, GDPval-AA v2.1, AA-Briefcase v1.1, AA-LCR v1.1 · Higher is better

Intelligence Evaluations

Intelligence evaluations measured independently by Artificial Analysis · Higher is better
查看更多

Agentic knowledge work, (Elo-500)/2000

Agentic SaaS workflows

Agentic real-world work tasks, (Elo-500)/2000

Reasoning & knowledge

Professional document reasoning, All-pass

Agentic coding & terminal use

Long context reasoning

SciCodeUnder review

Coding

CritPtUnder review

Physics reasoning

Agentic scientific research workflows in a terminal

Agentic software engineering in open-source repositories

Quantitative analysis on spreadsheets & documents

Legal agentic work, Hallucination-Gated All-Pass Rate

Kubernetes incident root-cause analysis

Medical long context reasoning

Visual reasoning

DeepSeek 的更多发布

更多详情

权重
服务商基准测试
DeepSeek V4.1 Flash (Max)
DeepSeek 标志DeepSeek
39
552B
推理时启用 16B 个参数
1M
US$0.2
179
NebiusLithosAICoreWeave
+15
DeepSeek V4.1 Flash (Non-reasoning)
DeepSeek 标志DeepSeek
25
552B
推理时启用 16B 个参数
1M
US$0.2
171
DeepSeekModalLithosAI
+4