Skip to main content
Back to Explore

Mistral: Mistral Large 3 2512

Mistral·Released 2025-12-01
Open Source262K ctxMoEMultimodal
CompareTry model

About

Mistral Large 3 2512 is Mistral’s most capable model to date, featuring a sparse mixture-of-experts architecture with 41B active parameters (675B total), and released under the Apache 2.0 license.

Related Models

Mistral: Mistral Medium 3.52026-04-30Mistral Small 4 (Reasoning)2026-03-16Mistral Small 4 (Non-reasoning)2026-03-16Mistral: Mistral Small 42026-03-16Mistral: Mistral Small Creative2025-12-16Mistral Large 32025-12-02Mistral: Mistral Medium 3.12025-08-13Mistral Small 3.22025-06-20

Pricing

Input

$0.50

per 1M tokens

Output

$1.50

per 1M tokens

Blended

$0.75

per 1M tokens

Cheaper than 44% of models. Median price is $0.54/1M tokens.

Cost Calculator

Tokens per day1M
100K100M

Daily

$0.75

Monthly

$22.50

vs. Similar Models

Mistral Large 3
$0.750%
Magistral Small 1.2
$0.750%
GPT-3.5 Turbo
$0.750%
Command-R (Mar '24)
$0.750%

Performance

Context Window

262K

tokens

Larger than 62% of models

Context Window Comparison

MoonshotAI: Kimi K2.6
262KSame
MoonshotAI: Kimi K2.7 Code
262KSame
Nex AGI: Nex-N2-Pro
262KSame

Open Source

Quick Compare

Similar Models

Mistral: Mistral Medium 3.5

Mistral

Q: 29.9$3.00/1M262K ctx
Pricier: 300%

Mistral Small 4 (Reasoning)

Mistral

Q: 20.8$0.26/1M
Cheaper: 65%

Devstral 2

Mistral

Q: 19.2N/A/1M

Magistral Medium 1.2

Mistral

Q: 17.9$2.75/1M
Pricier: 267%

Mistral Large 3

Mistral

Q: 15.9$0.75/1M

Mistral: Mistral Medium 3.1

Mistral

Q: 14.8$0.80/1M131K ctx
Context Window: 2x smaller
Compare all 7 models

Used by Agents

Gobii
Gobii