DeepSeek:模型智能、性能与价格

DeepSeek
DeepSeek

本分析旨在帮助你根据使用场景,选择 DeepSeek 提供的最佳模型。

最智能

#1
DeepSeek V4.1 Flash 1017 (max)DeepSeek V4.1 Flash 1017 (max)
33
#2
DeepSeek V4.1 Flash 1017 (high)DeepSeek V4.1 Flash 1017 (high)
29
#3
DeepSeek V4 Pro 0813 (max)DeepSeek V4 Pro 0813 (max)
23
#4
DeepSeek V4.1 Flash 1017 (non-reasoning)DeepSeek V4.1 Flash 1017 (non-reasoning)
19
#5
DeepSeek V4 Pro 0813 (non-reasoning)DeepSeek V4 Pro 0813 (non-reasoning)
13

Intelligence Index

共 5 个模型

每任务耗时最短

#1
DeepSeek V4.1 Flash 1017 (non-reasoning)DeepSeek V4.1 Flash 1017 (non-reasoning)
7.0 min
#2
DeepSeek V4.1 Flash 1017 (high)DeepSeek V4.1 Flash 1017 (high)
10.3 min
#3
DeepSeek V4.1 Flash 1017 (max)DeepSeek V4.1 Flash 1017 (max)
13.1 min
#4
DeepSeek V4 Pro 0813 (non-reasoning)DeepSeek V4 Pro 0813 (non-reasoning)
13.8 min
#5
DeepSeek V4 Pro 0813 (max)DeepSeek V4 Pro 0813 (max)
17.3 min

每个 Intelligence Index 任务耗时(分钟)

共 5 个模型

价格最低

#1
DeepSeek V4.1 Flash 1017 (max)DeepSeek V4.1 Flash 1017 (max)
$0.18
#2
DeepSeek V4.1 Flash 1017 (high)DeepSeek V4.1 Flash 1017 (high)
$0.18
#3
DeepSeek V4.1 Flash 1017 (non-reasoning)DeepSeek V4.1 Flash 1017 (non-reasoning)
$0.18
#4
DeepSeek V4 Pro 0813 (max)DeepSeek V4 Pro 0813 (max)
$0.69
#5
DeepSeek V4 Pro 0813 (non-reasoning)DeepSeek V4 Pro 0813 (non-reasoning)
$0.69

每 100 万 token 的混合价格

共 5 个模型

DeepSeek 提供 5 个模型,每个模型的智能、性能和价格特征各不相同。 下方对比了各模型的关键指标。

  • 智能方面,DeepSeek 上表现最好的模型是 DeepSeek V4.1 Flash 1017 (max)(33)、DeepSeek V4.1 Flash 1017 (high)(29)和DeepSeek V4 Pro 0813 (max)(23)。
  • 每任务耗时方面,DeepSeek V4.1 Flash 1017 (non-reasoning)(7.0 分钟)、DeepSeek V4.1 Flash 1017 (high)(10.3 分钟)和DeepSeek V4.1 Flash 1017 (max)(13.1 分钟) 在 DeepSeek 上完成一个 Intelligence Index 任务最快。 各模型之间的每任务耗时差异显著,最慢的模型耗时是最快模型的 2.5 倍。
  • 延迟方面,DeepSeek V4.1 Flash 1017 (non-reasoning)(1.26 秒)、DeepSeek V4 Pro 0813 (non-reasoning)(2.44 秒)和DeepSeek V4.1 Flash 1017 (high)(19.28 秒) 的首个答案 Token 延迟最低。
  • 价格方面,DeepSeek V4.1 Flash 1017 (max)($0.18)、DeepSeek V4.1 Flash 1017 (high)($0.18)和DeepSeek V4.1 Flash 1017 (non-reasoning)($0.18) 每 100 万 token 的混合价格最低。 各模型价格最多相差 3.8 倍。
  • 上下文窗口方面,DeepSeek V4 Pro 0813 (max)(1M)、DeepSeek V4 Pro 0813 (non-reasoning)(1M)和DeepSeek V4.1 Flash 1017 (max)(1M) 支持 DeepSeek 上最大的上下文窗口。
  • DeepSeek V4.1 Flash 1017 (max) 在智能与成本效益之间取得了最佳平衡。若追求最短的每任务耗时,DeepSeek V4.1 Flash 1017 (non-reasoning) 是首选。
Artificial Analysis Intelligence Index · Higher is better
Inference time (minutes) per Intelligence Index task · Lower is better
USD per 1M tokens (blended) · Lower is better

智能

Artificial Analysis Intelligence Index

Artificial Analysis Intelligence Index v5.0.1 incorporates 9 evaluations: AA-Briefcase v1.1, AutomationBench-AA, Terminal-Bench 4.0, FrontierCode-AA 2.0, AA-Omniscience v1.1, GDP.pdf, HLE-Diamond, CritPt v2 (Hard Subset), and Terminal-Bench-Science 0.1

Evaluations

Evaluations measured independently by Artificial Analysis · AA-PersonalityBench scores are percentile ranks from 1 to 100
查看更多

Agentic knowledge work, Elo

Agentic SaaS workflows

Agentic scientific research workflows in a terminal

Agentic software engineering in open-source repositories

Reasoning & knowledge

Professional document reasoning, All-pass

Agentic coding & terminal use

Physics reasoning

Quantitative analysis on spreadsheets & documents

No data available

Legal agentic work, Hallucination-Gated All-Pass Rate

No data available

Agentic business operations

Agentic real-world work tasks, Elo

Agentic interactive 3D experiences, Elo

No data available

Product and brand videos designed in code, Elo

No data available

Kubernetes incident root-cause analysis

No data available

Medical long context reasoning

Visual reasoning

Rich, varied, well-crafted language

Well-structured, scannable, answer-first

Truthful, acknowledges uncertainty, non-sycophantic

Use of filler and generic language

Intelligence Index 与价格

Blended at 7:2:1 (cache-input-output) · USD per 1M tokens (blended)
Most attractive quadrant

价格

每项 Intelligence Index 任务的成本

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

Intelligence Index 与价格

Blended at 7:2:1 (cache-input-output) · USD per 1M tokens (blended)
Most attractive quadrant

速度

已测量的单次请求和多轮智能体工作负载的平均值

Intelligence Index 每项任务耗时

Output tokens divided by end-to-end output speed (minutes per Intelligence Index task) · Lower is better

质量

Endpoint Accuracy Index

各接入点准确率相对其模型基准接入点的百分比 · 颜色表示差距是否具有统计显著性

延迟

已测量的单次请求和多轮智能体工作负载的平均值

首个答案 Token 延迟

经过推理后首个答案 token 到达所需的秒数 · 已测量的单次请求和多轮智能体工作负载的平均值 · 越低越好

可靠性

在所有已测量的单次请求和多轮智能体工作负载中,接入点失败的请求占比

接入点错误率

接入点因错误、限速或超时而失败的请求占比 · 涵盖所有已测量的单次请求和多轮智能体工作负载 · 越低越好

缓存

缓存命中率

多轮智能体会话各次运行在最近四周的中位数,更新于 2026年10月26日 · 越高越好

每任务成本与缓存命中率

Weighted average cost (USD) per Intelligence Index task · Cache hit rate: median of the last four weeks, updated Oct 26, 2026
Most attractive quadrant
Pareto line

进一步分析
DeepSeek 标志
DeepSeek V4.1 Flash 1017 (max)
33
$0.60
13.9
44.38
0.0%
1M
开放
DeepSeek 标志
DeepSeek V4.1 Flash 1017 (high)
29
$0.46
10.9
21.48
0.0%
1M
开放
DeepSeek 标志
DeepSeek V4 Pro 0813 (max)
23
$1.77
19.2
55.28
0.0%
1M
开放
DeepSeek 标志
DeepSeek V4 Pro 0813 (high)
21
$0.85
15.4
57.79
0.0%
1M
开放
DeepSeek 标志
DeepSeek V4.1 Flash 1017 (non-reasoning)
19
$0.43
7.5
1.34
0.7%
1M
开放
DeepSeek 标志
DeepSeek V4 Pro 0813 (non-reasoning)
13
$0.98
14.2
2.48
0.0%
1M
开放
DeepSeek 标志
DeepSeek V3.2 Exp (non-reasoning)
--
--
--
--
--
128k
开放
DeepSeek 标志
DeepSeek V3.2 (non-reasoning)
--
--
--
--
--
128k
开放
DeepSeek 标志
DeepSeek V3.2 Exp
--
--
--
--
--
128k
开放
DeepSeek 标志
DeepSeek V3.2
--
--
--
--
--
128k
开放

关键定义

常见问题

关于 DeepSeek 的常见问题