This model is deprecated. We only continue performance benchmarking for the default 10k input token workload. Results for other workloads are historical and no longer updated.

Mistral has launched a newer model, Mistral Small 4. We suggest considering it instead.

For more information, see comparison of Mistral Small 4 to other models and API provider benchmarks for Mistral Small 4.

Mistral Small 3.2 logo
Mistral
•

Open weights model

•

Released June 2025

Mistral Small 3.2 API 服务商基准测试与分析

模型比较

分析 Mistral Small 3.2 各 API 服务商的性能指标，包括延迟（首 Token 延迟）、输出速度（每秒输出 token 数）、价格等。参与基准测试的 API 服务商包括 DeepInfra (FP8)和Mistral。

速度最快

#1
MistralMistral
132.1 t/s
#2
DeepInfra (FP8)DeepInfra (FP8)
28.0 t/s

输出速度

共 2 家服务商

延迟最低

#1
DeepInfra (FP8)DeepInfra (FP8)
0.89 s
#2
MistralMistral
1.00 s

首 Token 延迟

共 2 家服务商

价格最低

#1
DeepInfra (FP8)DeepInfra (FP8)
$0.09
#2
MistralMistral
$0.12

每 100 万 token 的混合价格

共 2 家服务商

可通过 2 家 API 服务商使用 Mistral Small 3.2，每家服务商的性能特征和价格各不相同。下方对比了各服务商的关键指标。

  • 输出速度最快的服务商是 Mistral（132.1 t/s）和DeepInfra (FP8)（28.0 t/s）。
  • 延迟方面，DeepInfra (FP8)（0.89 秒）和Mistral（1.00 秒） 的首 Token 延迟最低。
  • 价格方面，DeepInfra (FP8)（0.09）和Mistral（0.12） 每 100 万 token 的混合价格最低。
  • DeepInfra (FP8) 同时拥有最低延迟和最佳价格，对注重成本的应用很有吸引力。对于吞吐量密集型工作负载，Mistral 速度最快。

亮点

速度

Output tokens per second · Higher is better

端到端响应时间

Seconds · Lower is better

价格

USD per 1M tokens (blended) · Lower is better

价格

Pricing: Cache Hit, Input, and Output

Price (USD per M Tokens) · Lower is better · 10,000 input tokens

Price per token for cached prompts (previously processed), typically offering a significant discount compared to regular input price, represented as USD per million tokens. The values shown here are the cache hit price; cache write and cache storage are billed separately and vary by provider — see "Cache pricing by provider" for detail.

Pricing: Blended Price

Blended at 7:2:1 (cache-input-output) · USD per 1M tokens (blended) · Lower is better

Price per token, shown in USD per million tokens. Price is a blend of cache hit, input, and output token prices using the selected ratio (default 7:2:1 cache-input-output).

Output Speed vs. Price

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

Tokens per second received while the model is generating tokens (ie. after first chunk has been received from the API for models which support streaming).

速度

按输出速度（每秒 token 数）衡量

Output Speed: Mistral Small 3.2

Output speed: output tokens per second · 10,000 input tokens

Tokens per second received while the model is generating tokens (ie. after first chunk has been received from the API for models which support streaming).

Figures represent performance of the model's first-party API (e.g. OpenAI for o1) or the median across providers where a first-party API is not available (e.g. Meta's Llama models).

Latency vs. Output Speed

Latency: seconds to first token received · Output speed: output tokens per second · 10,000 input tokens
Most attractive quadrant
Pareto line

Tokens per second received while the model is generating tokens (ie. after first chunk has been received from the API for models which support streaming).

延迟

按首 Token 延迟（秒）衡量

首 Token 延迟

Seconds to first token received · Lower is better · 10,000 input tokens

Time to first token received, in seconds, after API request sent. For reasoning models which share reasoning tokens, this will be the first reasoning token. For models which do not support streaming, this represents time to receive the completion.

Figures represent median (P50) measurement over the past 72 hours to reflect sustained changes in performance.

端到端响应时间

Seconds to output 500 tokens, calculated based on time to first token, 'thinking' time for reasoning models, and output speed

端到端响应时间

Seconds to output 500 tokens, including reasoning model 'thinking' time · Lower is better · 10,000 input tokens

Seconds to receive a 500 token response. Key components:

  • Input time: Time to receive the first response token
  • Thinking time (only for reasoning models): Time reasoning models spend outputting tokens to reason prior to providing an answer. Amount of tokens based on the average reasoning tokens across a diverse set of 60 prompts (methodology details).
  • Answer time: Time to generate 500 output tokens, based on output speed

关键比较指标与 API 功能

DeepInfra 标志DeepInfra
128k
开放
$0.10
28
0.87
18.70
--
Mistral 标志Mistral
131k
开放
$0.14
134
0.93
4.65
--

常见问题

关于 Mistral Small 3.2 服务商的常见问题

可通过 2 家 API 服务商使用 Mistral Small 3.2：DeepInfra (FP8)和Mistral。每家服务商的性能特征和价格各不相同。

目前可通过我们进行基准测试和跟踪的 2 家 API 服务商使用 Mistral Small 3.2。

按输出速度计算，Mistral Small 3.2 速度最快的服务商是 Mistral（132.1 t/s）和DeepInfra (FP8)（28.0 t/s）。输出速度衡量模型开始响应后生成 token 的速度。

Mistral Small 3.2 的首 Token 延迟最低的服务商是 DeepInfra (FP8)（0.89s）和Mistral（1.00s）。延迟越低，初始响应越快。

按混合价格计算，Mistral Small 3.2 最实惠的服务商是 DeepInfra (FP8)（每 100 万 token $0.09）和Mistral（每 100 万 token $0.12）。混合价格采用缓存命中/输入/输出 token 为 7:2:1 的比例。

Mistral Small 3.2 输入 token 价格最低的服务商是 DeepInfra (FP8)（每 100 万输入 token $0.07）和Mistral（每 100 万输入 token $0.10）。

Mistral Small 3.2 输出 token 价格最低的服务商是 DeepInfra (FP8)（每 100 万输出 token $0.20）和Mistral（每 100 万输出 token $0.30）。

Mistral Small 3.2 各服务商价格最多相差 1.4 倍。最实惠的是 DeepInfra (FP8)，每 100 万 token 收费 $0.09；Mistral 则收费 $0.12。

Mistral Small 3.2 的输出速度在不同服务商之间差异显著。Mistral 最快，为 132.1 t/s，是 DeepInfra (FP8)（28.0 t/s）的 4.7 倍。

提供 Mistral Small 3.2 的全部 2 家服务商均支持用于结构化输出的 JSON 模式。

提供 Mistral Small 3.2 的全部 2 家服务商均支持函数调用（工具使用）。

Mistral Small 3.2 的最佳服务商取决于你的侧重点：Mistral 输出速度最高，DeepInfra (FP8) 延迟最低，DeepInfra (FP8) 的价格最具竞争力。

选择 Mistral Small 3.2 服务商时，请考虑：输出速度（适合吞吐量密集型任务）、延迟（适合需要快速首次响应的交互式应用）、价格（适合成本敏感型工作负载），以及 JSON 模式或函数调用等 API 功能。

会。服务商性能可能因基础设施变化、负载均衡和更新而随时间波动。我们持续对所有服务商进行基准测试，并在“随时间变化”图表中展示历史性能趋势。

有关 Mistral Small 3.2 的智能、能力、模态及其与其他模型的比较，请参阅模型概览页面。 查看模型概览