Mistral:模型智能、性能与价格

Mistral
Mistral

本分析旨在帮助你根据使用场景,选择 Mistral 提供的最佳模型。

最智能

#1
Mistral Large 4 (high)Mistral Large 4 (high)
26
#2
Mistral Large 4 PreviewMistral Large 4 Preview
25
#3
Mistral Large 4 (medium)Mistral Large 4 (medium)
23
#4
Mistral Medium 3.5Mistral Medium 3.5
5
#5
Mistral Small 4Mistral Small 4
4

Intelligence Index

共 11 个模型

每任务耗时最短

#1
Mistral Small 4Mistral Small 4
1.7 min
#2
Mistral Medium 3.5Mistral Medium 3.5
4.5 min
#3
Mistral Large 4 (medium)Mistral Large 4 (medium)
12.5 min
#4
Mistral Large 4 (high)Mistral Large 4 (high)
18.4 min
#5
Mistral Large 4 PreviewMistral Large 4 Preview
26.0 min

每个 Intelligence Index 任务耗时(分钟)

共 11 个模型

价格最低

#1
Ministral 3 3BMinistral 3 3B
$0.04
#2
Ministral 3 8BMinistral 3 8B
$0.06
#3
Ministral 3 14BMinistral 3 14B
$0.07
#4
Mistral Small 4Mistral Small 4
$0.10
#5
Mistral Small 4 (non-reasoning)Mistral Small 4 (non-reasoning)
$0.10

每 100 万 token 的混合价格

共 11 个模型

Mistral 提供 11 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,Mistral 上表现最好的模型是 Mistral Large 4 (high)(26)、Mistral Large 4 Preview(25)和Mistral Large 4 (medium)(23)。
  • 每任务耗时方面,Mistral Small 4(1.7 分钟)、Mistral Medium 3.5(4.5 分钟)和Mistral Large 4 (medium)(12.5 分钟) 在 Mistral 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 15.3 倍。
  • 延迟方面,Mistral Small 4 (non-reasoning)(1.62 秒)、Ministral 3 3B(1.67 秒)和Ministral 3 8B(1.84 秒) 的首个答案 Token 延迟最低。
  • 价格方面,Ministral 3 3B($0.04)、Ministral 3 8B($0.06)和Ministral 3 14B($0.07) 每 100 万 token 的混合价格最低。 各模型价格最多相差 2.7 倍。
  • 上下文窗口方面,GLM-5.2 (max)(1M)、Mistral Large 4 (high)(1M)和Mistral Large 4 (medium)(1M) 支持 Mistral 上最大的上下文窗口。
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

Legal agentic work, Hallucination-Gated All-Pass Rate

Agentic business operations

Agentic real-world work tasks, Elo

Agentic interactive 3D experiences, Elo

Product and brand videos designed in code, Elo

Kubernetes incident root-cause analysis

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
Pareto line

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
Pareto line

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

上下文窗口

Context window: tokens limit · Higher is better

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
Mistral 标志
Mistral Large 4 (high)
26
$2.06
16.7
55.82
0.0%
1M
开放
Mistral 标志
Mistral Large 4 Preview
25
$4.21
23.5
40.64
0.0%
524k
专有
Mistral 标志
Mistral Large 4 (medium)
23
$1.41
12.1
35.24
0.0%
1M
开放
Mistral 标志
Mistral Medium 3.5
5
$1.04
4.6
23.97
0.0%
262k
开放
Mistral 标志
Mistral Small 4
4
$0.0399
1.7
12.58
1.7%
256k
开放
Mistral 标志
Mistral Small 4 (non-reasoning)
--
--
--
1.61
5.7%
262k
开放
Mistral 标志
Mistral Large 3
--
--
--
2.65
2.2%
262k
开放
Mistral 标志
Ministral 3 3B
--
--
--
1.69
4.6%
131k
开放
Mistral 标志
Ministral 3 14B
--
--
--
2.53
3.4%
262k
开放
Mistral 标志
Ministral 3 8B
--
--
--
1.84
2.5%
262k
开放
Z AI 标志
GLM-5.2 (max)
--
--
--
8.62
3.9%
1M
开放

关键定义

常见问题

关于 Mistral 的常见问题