Passer au contenu principal
AITier

Mistral Medium 3.5

MistralMistralOpen WeightModified MIT License

Description

Mistral Medium 3.5 is Mistral AI's first flagship merged model: a dense 128B-parameter multimodal model with a 256k context window that unifies instruction-following, reasoning, and coding capabilities in a single set of weights. It replaces Mistral Medium 3.1 and Magistral in Le Chat, and Devstral 2 in the Vibe coding agent. Reasoning effort is configurable per request, and the vision encoder was trained from scratch to handle variable image sizes and aspect ratios. Released as open weights under a Modified MIT License.

Date de sortie
2026-04-29
Paramètres
128.0B
Longueur du contexte
262K
Modalités
image, text

Radar de capacités

14
general
46
coding
75
reasoning
50
science
10
agents
75
multimodal

Classements

Domaine#RangScoreSource
Capacité agentique99
35.0
LS
Classement codage234
61.0
AA
Classement général203
54.0
AA
Science266
50.0
AA

Scores de benchmarks (LLM Stats)

(LLM Stats (zeroeval))

Agents

Tau3 Telecom91.4%Aut.
Tau3 Retail76.1%Aut.
Tau3 Airline72.0%Aut.
Finance Agent v232.1%
Tau3 Banking13.4%Aut.

Instruction Following

IFBench69.0%Aut.

Language

COLLIE95.8%Aut.

Math

AIME 202586.3%Aut.
Beyond AIME66.9%Aut.

Reasoning

SWE-Bench Verified77.6%Aut.
BrowseCompOpenAI (2025)48.6%Aut.

Indices d'évaluation AA

(Artificial Analysis)
Tau2(Sierra + U Toronto + Vector Institute (2025))
94.2
Gpqa(NYU + Cohere + Anthropic (2023))
74.8
Lcr(Artificial Analysis)
69.3
Ifbench(Google Research (2023))
68.8
Terminalbench V2 1
50.6
Coding Index(Artificial Analysis)
46.9
Scicode(UIUC + Argonne National Lab (2024))
40.2
Terminalbench Hard(Stanford × Laude Institute (2026))
33.3
Tau Banking
15.1
Intelligence Index(Artificial Analysis)
14.2
Hle(Center for AI Safety + Scale AI (2025))
13.8
Terminalbench V4 0
0.0

Scores par catégorie LLM Stats

(LLM Stats (zeroeval))
Language
100
Writing
100
Math
80
Frontend Development
80
Code
80
Instruction Following
70
Reasoning
60
General
60
Search
50
Finance
30
Agents
30
Tool Calling
10

Tarification

Prix d'entrée$1.5 / 1M tokens
Prix de sortie$7.5 / 1M tokens
Prix mixte (3:1)$3 / 1M tokens
Prix de lecture cache$0.15 / 1M tokens

Vitesse

Tokens/sec168.1
Délai du premier token0.61s
Temps de réponse12.51s

Classement des Prix par Fournisseur

Classement des Prix par Fournisseur

6 fournisseurs

Moins cher: Mistral AIPlus cher: Requesty
FournisseurEntréeSortie
1Mistral AIMoins cher
$0
$0.00001
2MistralPRINCIPAL
$1.5
$7.5
3OpenRouter
$1.5
$7.5
4Kilo Gateway
$1.5
$7.5
5DevPass (LLM Gateway)
$1.5
$7.5
6Requesty
$1.65
$8.25

Comparer les prix entre différents fournisseurs API pour ce modèle.

Sources externes

LLM StatsArtificial Analysis