Saltar al contenido principal
AITier

Ministral 3 14B

MistralMistralOpen WeightApache 2.0 · Uso Comercial

Descripción

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 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 14B can even be deployed locally, capable of fitting in 24GB of VRAM in BF16, and less than 12GB of RAM/VRAM when quantized.

Fecha de lanzamiento
2025-12-02
Parámetros
14.0B
Longitud del contexto
262K
Modalidades
image, text

Radar de capacidades

25
general
23
coding
35
reasoning
35
science
31
agents
10
multimodal

Rankings

Dominio#PosiciónPuntuaciónFuente
Ranking de codificación494
22.0
AA
Ranking general512
28.0
AA
Ciencia474
30.0
AA

Puntuaciones de benchmarks (LLM Stats)

(LLM Stats (zeroeval))

Chat

MM-MT-Bench0.08 / 100Aut.

General

MMLU79.4%Aut.
TriviaQA74.9%Aut.
Multilingual MMLU74.2%Aut.
Wild Bench68.5%Aut.
Arena Hard55.1%Aut.

Language

MMLU-Redux82.0%Aut.

Math

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

Reasoning

GPQANYU + Cohere + Anthropic (2023)71.2%Aut.
AGIEval64.8%Aut.
LiveCodeBench64.6%Aut.

Índices de evaluación AA

(Artificial Analysis)
Mmlu Pro(TIGER-Lab (Univ. of Waterloo, Toronto, CMU, 2024))
69.3
Gpqa(NYU + Cohere + Anthropic (2023))
57.2
Livecodebench(UC Berkeley + MIT + Cornell (2024))
35.1
Ifbench(Google Research (2023))
32.0
Math Index(Artificial Analysis)
30.0
Aime 25(MAA (Mathematical Association of America))
30.0
Tau2(Sierra + U Toronto + Vector Institute (2025))
27.2
Lcr(Artificial Analysis)
26.3
Scicode(UIUC + Argonne National Lab (2024))
23.8
Coding Index(Artificial Analysis)
14.4
Terminalbench V2 1
9.7
Tau Banking
6.6
Intelligence Index(Artificial Analysis)
6.0
Hle(Center for AI Safety + Scale AI (2025))
4.6
Terminalbench Hard(Stanford × Laude Institute (2026))
4.5
Terminalbench V4 0
0.0

Puntuaciones por categoría LLM Stats

(LLM Stats (zeroeval))
Math
90
Reasoning
70
General
60
Creativity
60
Writing
60
Communication
40
Chat
30
Multimodal
10

Precios

Precio de entrada$0.2 / 1M tokens
Precio de salida$0.2 / 1M tokens
Precio mixto (3:1)$0.2 / 1M tokens
Precio de lectura caché$0.02 / 1M tokens

Velocidad

Tokens/seg82.7
Retraso del primer token0.55s
Tiempo hasta la respuesta0.55s

Ranking de Precios por Proveedor

Ranking de Precios por Proveedor

4 proveedores

Más barato: MistralMás caro: Kilo Gateway
ProveedorEntradaSalida
1MistralPRINCIPAL
$0.2
$0.2
2NanoGPT
$0.2
$0.2
3OpenRouter
$0.2
$0.2
4Kilo Gateway
$0.2
$0.2

Comparar precios entre diferentes proveedores de API para este modelo.

Fuentes externas

LLM StatsArtificial Analysis