Mistral Large 4 — Specifications
Rédigé par Mustafa Ihsan d’après la documentation officielle du fournisseur · Dernière mise à jour
| Développeur | Mistral AI |
|---|---|
| Type | LLM (architecture MoE, raisonnement, multimodal) |
| Modalité | Texte, image → texte |
| Paramètres | 1.05T total / 52B active (MoE) |
| Fenêtre de contexte | 1 million |
| Licence | Mistral Ecosystem Open Weight License |
| Poids ouverts | Oui |
| Publié | 2026-10 |
| Prix de l’entrée | $0.70 /1M |
| Prix de la sortie | $2.10 /1M |
| Fournisseurs d'API | Mistral API, OpenRouter |
Exécutez-le localement
| VRAM (4 bits) | ~610 GB |
|---|
Résultats de référence
| Mistral-reported | Artificial Analysis Coding Agent Index 48%; Terminal-Bench 4.0 41%; Artificial Analysis Intelligence Index v5 26. |
|---|
What is Mistral Large 4?
Mistral Large 4 is Mistral AI’s largest and most capable model to date: a mixture-of-experts model with 1.05T total parameters, 52B active per token and a 1.6B vision encoder. It reads text and images, writes text, has a 1M-token context window and offers three reasoning efforts (low, medium and high, the default). Mistral opened a préversion publique on 6 October 2026 and released the weights of an improved version on 29 October 2026; its model page now marks it generally available.
Sur Mistral’s pricing page it costs $0.70 per million input tokens, $0.07 for cached input and $2.10 per million output tokens. The licence is the Mistral Ecosystem Open-Weight License, which Mistral describes as a modified MIT licence: you can use, modify, fine-tune and commercialise the model, but a business that runs a Model-as-a-Service offering needs a separate commercial licence from Mistral. Besides Mistral’s own API (mistral-large-4), Mistral says it is available on OpenRouter, through Microsoft Foundry and on NVIDIA-hosted endpoints.
Self-hosting is data-centre work. The FP8 checkpoint on Hugging Face totals about 1.06 TB of weights, and Mistral’s NVFP4 checkpoint, about 610 GB, targets Blackwell GPUs (B200, B300, GB200 and GB300). It succeeds Mistral Large 3, which Mistral deprecated on 31 October 2026 and plans to retire on 30 April 2027; Large 3 remains the option if you need Apache 2.0 terms. Weigh renting against owning with the calculatrice auto-hébergement vs API, or estimate a bill with the Calculateur de coûts pour les API IA.
Mistral Large 4 pricing: API cost per 1M tokens
| Entrée (par million de jetons) | $0.700 |
|---|---|
| Sortie (par million de jetons) | $2.10 |
| Ratio sortie/entrée | 3× |
| Moyenne pondérée (4:1 entrée:sortie) | $0.980 par 1 million de jetons |
What Mistral Large 4 costs per month
Dépense mensuelle réelle pour un ratio entrée-sortie de 4:1 — le rapport effectivement observé dans une charge de travail type (chat ou RAG).
| Charge de travail | Jetons/mois | Coût par mois |
|---|---|---|
| Projet secondaire | 1 million en entrée / 0,25 million en sortie | $1.23 |
| Petite équipe | 20 millions en entrée / 5 millions en sortie | $25 |
| Production | 200 millions en entrée / 50 millions en sortie | $245 |
Calculez vos propres coûts dans le Calculateur de coûts pour les API IA.
Cheaper alternatives to Mistral Large 4
| Modèle | Coût combiné par million de dollars | Vous économisez |
|---|---|---|
| Llama 3.1 8B ouverte | $0.0240 | 98% cheaper |
| Mistral NeMo 12B ouverte | $0.0292 | 97 % moins cher |
| Gemma 3 4B ouverte | $0.0600 | 94% cheaper |
Hébergement local ou paiement via API ?
Mistral Large 4 is open-weight, so you can run it yourself. It needs ~610 GB of VRAM at 4-bit. Self-hosting only beats the API once your volume is high enough to keep that hardware busy — the calculatrice auto-hébergement vs API calcule le seuil de rentabilité pour votre volume de jetons.
Questions fréquemment posées
How much does Mistral Large 4 cost per 1M tokens?
Mistral Large 4 costs $0.700 per 1M input tokens and $2.10 per 1M output tokens. At a typical 4:1 input-to-output mix that blends to about $0.980 per 1M tokens.
How much does Mistral Large 4 cost per month?
A small-team workload of 20M input and 5M output tokens a month costs about $25 on Mistral Large 4. A side project (1M in / 0.25M out) costs roughly $1.23.
What is a cheaper alternative to Mistral Large 4?
Llama 3.1 8B is the strongest cheaper option in our database at $0.0240 per 1M blended — about 98% less than Mistral Large 4. It is also open-weight, so self-hosting is an option.
Can I run Mistral Large 4 locally?
Yes. Mistral Large 4 is open-weight and needs about ~610 GB of VRAM at 4-bit quantisation.
Why does Mistral Large 4 charge more for output than input?
Output tokens are generated one at a time and cannot be batched the way a prompt can, so they cost the provider more to serve. Mistral Large 4 charges 3× more for output, which is why prompt-heavy workloads are far cheaper to run than generation-heavy ones.
Les prix indiqués correspondent aux tarifs publics officiels pour l’API principale du modèle et sont mis à jour régulièrement à mesure que les fournisseurs les modifient. Les remises liées aux volumes, aux traitements par lots ou aux entrées mises en cache ne sont pas prises en compte. Comparez tous les modèles côte à côte dans le Base de données des modèles IA ou le Classement des grands modèles linguistiques (LLM).