Ministral 3 3B
MistralMistral開源權重Apache 2.0 · 商用許可
描述
The smallest model in the Ministral 3 family, Ministral 3 3B is a powerful, efficient tiny language model with vision capabilities. This model is the instruct post-trained version in FP8, fine-tuned for instruction tasks, making it ideal for chat and instruction based use cases. The Ministral 3 family is designed for edge deployment, capable of running on a wide range of hardware. Ministral 3 3B can even be deployed locally, fitting in 16GB of VRAM in BF16, and less than 8GB of RAM/VRAM when quantized.
發布日期
2025-12-02
參數規模
3.0B
上下文長度
131K
支援模態
image, text
能力雷達圖
19
general
13
coding
24
reasoning
23
science
21
agents
10
multimodal
排行榜排名
基準測試分數 (LLM Stats)
(LLM Stats (zeroeval))Chat
MM-MT-Bench
0.08 / 100自報
General
MMLU
70.7%自報
Multilingual MMLU
65.2%自報
TriviaQA
59.2%自報
Wild Bench
56.8%自報
Arena Hard
30.5%自報
Language
MMLU-Redux
73.5%自報
Math
MATH
83.0%自報
AIME 2024
77.5%自報
AIME 2025
72.1%自報
MATH (CoT)
60.1%自報
Reasoning
LiveCodeBench
54.8%自報
GPQANYU + Cohere + Anthropic (2023)
53.4%自報
AGIEval
51.1%自報
AA 評測指數
(Artificial Analysis)Mmlu Pro(TIGER-Lab (Univ. of Waterloo, Toronto, CMU, 2024))52.4
Gpqa(NYU + Cohere + Anthropic (2023))35.8
Ifbench(Google Research (2023))26.8
Tau2(Sierra + U Toronto + Vector Institute (2025))24.9
Livecodebench(UC Berkeley + MIT + Cornell (2024))24.7
Math Index(Artificial Analysis)22.0
Aime 25(MAA (Mathematical Association of America))22.0
Lcr(Artificial Analysis)17.0
Scicode(UIUC + Argonne National Lab (2024))15.3
Hle(Center for AI Safety + Scale AI (2025))5.4
Intelligence Index(Artificial Analysis)4.8
Coding Index(Artificial Analysis)4.8
Tau Banking4.7
Terminalbench Hard(Stanford × Laude Institute (2026))0.0
Terminalbench V2 10.0
Terminalbench V4 00.0
LLM Stats 分類評分
(LLM Stats (zeroeval))Math80
Reasoning60
General40
Communication30
Creativity30
Writing30
Chat20
Multimodal10
定價
輸入價格$0.1 / 1M tokens
輸出價格$0.1 / 1M tokens
混合價格(3:1)$0.1 / 1M tokens
快取讀取價格$0.01 / 1M tokens
速度
Tokens/秒242.8
首Token延遲0.58s
首回答延遲0.58s
供應商價格排行
供應商價格排行
4 個供應商
最便宜: Mistral最貴: Kilo Gateway
供應商輸入輸出
1Mistral主要
$0.1
$0.1
2NanoGPT
$0.1
$0.1
3OpenRouter
$0.1
$0.1
4Kilo Gateway
$0.1
$0.1
比較該模型在不同 API 供應商之間的定價。