跳转到主要内容

Ministral 3 14B

MistralMistral开源权重Apache 2.0 · 商用许可

描述

A balanced model in the Ministral 3 family, Ministral 3 14B is a powerful, efficient tiny language model with vision capabilities. This model is the reasoning post-trained version, trained for reasoning tasks, making it ideal for math, coding and stem related use cases. The Ministral 3 family is designed for edge deployment, capable of running on a wide range of hardware. Ministral 3 14B can even be deployed locally, capable of fitting in 24GB of VRAM in BF16, and less than 12GB of RAM/VRAM when quantized.

发布日期
2025-12-02
参数规模
14.0B
上下文长度
262K
支持模态
image, text

能力雷达图

28
general
23
coding
35
reasoning
35
science
31
agents
10
multimodal

排行榜排名

领域#排名分数来源
代码能力榜406
22.0
AA
通用能力榜384
33.0
AA
科学能力389
34.0
AA

基准测试分数 (LLM Stats)

(LLM Stats (zeroeval))

Biology

GPQANYU + Cohere + Anthropic (2023)71.2%自报

Code

LiveCodeBench64.6%自报

Communication

Wild Bench68.5%自报
MM-MT-Bench0.08 / 100自报

Creativity

Arena Hard55.1%自报

Finance

MMLU79.4%自报

General

MMLU-Redux82.0%自报
TriviaQA74.9%自报
Multilingual MMLU74.2%自报
AGIEval64.8%自报

Math

MATH90.4%自报
AIME 202489.8%自报
AIME 202585.0%自报
MATH (CoT)67.6%自报

AA 评测指数

(Artificial Analysis)
Math Index(Artificial Analysis)
30.0
Coding Index(Artificial Analysis)
14.4
Intelligence Index(Artificial Analysis)
11.2
Mmlu Pro(TIGER-Lab (Univ. of Waterloo, Toronto, CMU, 2024))
0.7
Gpqa(NYU + Cohere + Anthropic (2023))
0.6
Livecodebench(UC Berkeley + MIT + Cornell (2024))
0.4
Ifbench(Google Research (2023))
0.3
Aime 25(MAA (Mathematical Association of America))
0.3
Tau2(Sierra + U Toronto + Vector Institute (2025))
0.3
Lcr(Artificial Analysis)
0.3
Scicode(UIUC + Argonne National Lab (2024))
0.2
Terminalbench V2 1
0.1
Tau Banking
0.1
Hle(Center for AI Safety + Scale AI (2025))
0.0
Terminalbench Hard(Stanford × Laude Institute (2026))
0.0

LLM Stats 分类评分

(LLM Stats (zeroeval))
Math
90
Reasoning
80
General
80
Physics
70
Biology
70
Chemistry
70
Code
60

定价

输入价格$0.2 / 1M tokens
输出价格$0.2 / 1M tokens
混合价格(3:1)$0.2 / 1M tokens
缓存读取价格$0.02 / 1M tokens

速度

Tokens/秒93.9
首Token延迟0.53s
首回答延迟0.53s

供应商价格排行

供应商价格排行

4 个供应商

最便宜: Mistral最贵: Kilo Gateway
供应商输入输出
1Mistral主要
$0.2
$0.2
2NanoGPT
$0.2
$0.2
3OpenRouter
$0.2
$0.2
4Kilo Gateway
$0.2
$0.2

比较该模型在不同 API 供应商之间的定价。

外部链接