by Mistral AI

Mistral Small 4.

multimodal open weights datacenter 121B params 262K ctx
Cheapest input
$0.15/M
on Mistral La Plateforme
Cheapest output
$0.6/M
on Mistral La Plateforme
Smallest GPU
2× Nvidia Q RTX 8000
$0.54/hr

Mistral Small 4 is an Apache 2.0 mixture-of-experts model, 119B total and 6.5B active, with text and image input, 256k context and per-request reasoning.

Smallest GPU to run it See all quantisations →

2× Nvidia Q RTX 8000 · $0.54/hr.

Most-aggressive quantisation we have a working recommendation for. Lower precision = less VRAM = cheaper hardware, at a small accuracy cost.

Where to use it

Cheapest hosted endpoints.

Provider Access $/M in $/M out
Mistral La Plateforme api direct $0.15 $0.6 Launch ↗
OpenRouter api aggregator $0.15 $0.6 Launch ↗
Sources

Official references.

Family

Variants in the Mistral family.

Mistral 7B Base v0.1
open apache-2.0
Mistral Large 2
123.0B open Mistral Research License

Mistral's flagship open-weight model — 123B dense.

Mistral 7B v0.3
7.3B open Apache-2.0

The current Mistral 7B — adds function calling + extended vocab.

Mistral 7B v0.2
7.2B open Apache-2.0

Mistral 7B v0.2 — earlier 32K context revision.

Mistral 7B v0.1
7.2B open Apache-2.0

Original Mistral 7B — historical reference.

Mistral Nemo 12B
12.2B open Apache-2.0

Mistral × Nvidia collab — 12B Apache-licensed, multilingual.

Mistral Large 4
open

Mistral Large 4 is a frontier multimodal (text and image input) model from Mi...

Mistral Large 3 2512
closed

Mistral Large 3 2512 is Mistral’s most capable model to date, featuring a spa...

Mistral Medium 3
closed

Mistral Medium 3 is a high-performance enterprise-grade language model design...

Mistral Medium 3.1
closed

Mistral Medium 3.1 is an updated version of Mistral Medium 3, which is a high...

Mistral Medium 3.5
closed

Mistral Medium 3.5 is a dense 128B open-weights model for instructions, reaso...

Mistral Small 3
24.0B open apache-2.0

Mistral Small 3 is a 24B-parameter language model optimized for low-latency p...

Mistral Small 3.1 24B
24.0B open apache-2.0

Mistral Small 3.1 24B Instruct is an upgraded variant of Mistral Small 3 (250...

Mistral Small 3.2 24B
24.0B open apache-2.0

Mistral-Small-3.2-24B-Instruct-2506 is an updated 24B parameter model from Mi...

Saba
closed

Mistral Saba is a 24B-parameter language model specifically designed for the ...

FAQ

Frequently asked.

How do I run Mistral Small 4?
Mistral Small 4 is open-weight, so you can self-host on rented GPUs. See the Run It Yourself tab for GPU configurations + cost estimates, or use one of the hosted inference providers listed on this page.
Where can I access Mistral Small 4?
Mistral Small 4 is available via Mistral La Plateforme, OpenRouter. Each access option lists its own pricing (per million tokens or hourly hosting).
How much does it cost to run Mistral Small 4?
API pricing starts at $0.15/M input tokens and $0.6/M output tokens. Self-hosting cost depends on the GPU you rent — see the Run It Yourself tab.
Is Mistral Small 4 open-source or proprietary?
Mistral Small 4 is open-weight under the apache-2.0 license. You can download and self-host it.