LLM ReferenceLLM Reference

Mistral Large

4 models2024–2025Up to 128K ctxFrom $0.32/1M input

About

Mistral AI's top-tier general-purpose family, covering the Large 2 and Large 3 product lines.

Specifications(4 models)

Mistral Large model specifications comparison
ModelReleasedContextParametersVisionMultimodalFn CallingTool UseStructured Outputs
Mistral Large 3 675B Instruct2025-12128K675BNoNoNoNoYes
Mistral Large 22025-11128K123BYesYesYesYesYes
Mistral Large 2 (2407)2024-07128K123BYesNoNoNoYes

Available From(10 providers)

Pricing

Mistral Large model pricing by provider
ModelProviderInput / 1MOutput / 1MType
Mistral Large 2AWS Bedrock$0.48$2.4Serverless
Mistral Large 2OpenRouter$0.5$1.5Serverless
Mistral Large 3 675B InstructAWS Bedrock$0.5$1.5Serverless
Mistral Large 2 (2407)Chutes AI$0.5$1.5Serverless
Mistral Large 2IBM watsonx$2$6Serverless
Mistral Large 2 (2407)SiliconFlow$2$2Serverless
Mistral Large 2 (2407)Microsoft Foundry$3$9Serverless

Frequently Asked Questions

What is Mistral Large?
Mistral AI's top-tier general-purpose family, covering the Large 2 and Large 3 product lines.
How many models are in the Mistral Large family?
The Mistral Large family contains 4 models.
What is the latest Mistral Large model?
The latest model is Mistral Large 3 675B Instruct, released in 2025-12.
How much does Mistral Large cost?
Mistral Large models range from $0.32/1M to $10/1M input tokens depending on the model and provider.
Is Mistral Large open source?
1 of 4 Mistral Large models are open source.

Models(4)