Mistral Large 3 at a glance
Mistral 3 includes a 675B-parameter mixture-of-experts flagship and dense 3B, 8B, and 14B models, all open-weight under Apache 2.0 with image understanding.
What changed with Mistral Large 3
On December 2, 2025, Mistral AI released Mistral 3. Mistral Large 3 is a sparse mixture-of-experts model with 41B active and 675B total parameters, and Ministral 3 comes in dense 3B, 8B, and 14B sizes. All are released under Apache 2.0, support image input, and handle many languages.
The models launched on Mistral AI Studio, Amazon Bedrock, Azure Foundry, Hugging Face, and other providers. Listed API prices are $0.50/$1.50 per 1M input/output tokens for Large 3 and $0.15/$0.15 for Ministral 3 8B.
Mistral Large 3 benchmarks and pricing
Values are copied from the linked source on the date shown. They are not IA Compa measurements.
- Input price per 1M tokensas of · Source for Input price per 1M tokens: mistral.ai (opens in a new tab)
- $0.50
- Output price per 1M tokensas of · Source for Output price per 1M tokens: mistral.ai (opens in a new tab)
- $1.50
How it compares
Other models in the same category that we have covered, using each model's latest recorded value. (Models)
Where to compare Mistral Large 3
Independent leaderboards and official resources for comparing this model yourself. Rankings change often; each site shows its own current data.
Leaderboards
General leaderboards
Why Mistral Large 3 matters for businesses
A permissive license on a large model matters for businesses that need to self-host for privacy or cost control, and the small Ministral models suit on-device or low-cost classification and extraction jobs.
Sources for this story
- Introducing Mistral 3 (opens in a new tab)Official
Mistral AI · mistral.ai
Last verified: . IA Compa summarizes public announcements for business readers. We are not affiliated with the companies mentioned. Always confirm details, pricing, and availability with the official source.
- #llm
- #mistral
- #open-weights
- #apache-2-0
