LLM Reference

Mistral Small Models by MistralAI

MistralAIApache 2.0Open source
5 models2024–2026Up to 256k ctxFrom $0.1/1M input

Details

ResearcherMistralAI
LicenseApache 2.0OSI-approved
Commercial useCommercial use: permitted
Models5
Released2024–2026
Max context256k

Capabilities

Vision3 of 5 models
Multimodal3 of 5 models
Function Calling3 of 5 models
Tool Use3 of 5 models
Structured Outputs4 of 5 models
Code Execution1 of 5 models

Links

Website

About

Mistral AI's small frontier model family for general-purpose chat, reasoning, and agent use cases.

Current Variants

Use-when guidance is based on each model's tracked capabilities, context window, release date, and replacement status.

5 in view

Use when the workload needs 256k context, tool use, and function calling.

2026-03256k contexttool usefunction calling

Use when the workload needs 128k context, 24B parameters, and structured outputs.

2025-12128k context24B parametersstructured outputs

Use when the workload needs 128k context, 24B parameters, and tool use.

2025-06128k context24B parameterstool use

Use when the workload needs 33k context, 24B parameters, and tool use.

2025-0133k context24B parameterstool use

Use when the workload needs 32k context, 22B parameters, and structured outputs.

2024-0232k context22B parametersstructured outputs

Release Timeline

5 release groups
2026-03
1 current
Mistral Small 4
256k contexttool usefunction calling
Current
2025-12
1 current
Mistral Small 3.1 24B Instruct
128k context24B parametersstructured outputs
Current
2025-06
1 current
Mistral Small 3.2 24B
128k context24B parameterstool use
Current
2025-01
1 current
Mistral Small 3
33k context24B parameterstool use
Current
2024-02
1 current
Mistral Small
32k context22B parametersstructured outputs
Current

Specifications(5 models)

Mistral Small model specifications comparison
ModelReleasedContextParametersVisionMultimodalFn CallingTool UseStructured OutputsCode Exec
Mistral Small 42026-03256k119B (6.5B active)YesYesYesYesNoNo
Mistral Small 3.1 24B Instruct2025-12128k24BYesYesNoNoYesNo
Mistral Small 3.2 24B2025-06128k24BYesYesYesYesYesYes
Mistral Small 32025-0133k24BNoNoYesYesYesNo
Mistral Small2024-0232k22BNoNoNoNoYesNo

Pricing

Mistral Small model pricing by provider
ModelProviderInput / 1MOutput / 1MType
Mistral SmallMistral AI Studio$0.1$0.3Serverless
Mistral Small 3Together AI$0.1$0.3Serverless
Mistral Small 3.1 24B InstructTogether AI$0.1$0.3Serverless
Mistral Small 4Mistral AI Studio$0.1$0.3Serverless
Mistral Small 3Vercel AI Gateway$0.1$0.3Serverless
Mistral Small 4OpenRouter$0.15$0.6Serverless
Mistral Small 3.1 24B InstructOpenRouter$0.35$0.56Serverless
Mistral Small 3.1 24B InstructCloudflare Workers AI$0.351$0.555Serverless
Mistral Small 3.1 24B InstructFireworks AI$0.9$0.9Serverless
Mistral SmallMicrosoft Foundry$1$3Provisioned
Mistral SmallAWS Bedrock$1$3Serverless

Popular comparisons in this family

Frequently Asked Questions

What is Mistral Small used for?
Mistral Small is used for vision and multimodal work, agent workflows and tool use, and structured outputs. The family description and listed model capabilities point to those workloads as the best fit.
How does Mistral Small compare to Ministral?
Mistral Small by MistralAI is strongest where you need vision and multimodal work, while Ministral by MistralAI is the closest related family to check for vision and multimodal work. Mistral Small has 5 listed variants and reaches up to 256k context, while Ministral reaches up to 32k context, so compare the specs and pricing tables before choosing a production model.
Which Mistral Small model should I use?
For the lowest listed input price, start with Mistral Small through Mistral AI Studio at $0.1/1M input tokens. For the most capable/latest local choice, evaluate Mistral Small 3.2 24B with 128k context and tool use, function calling, structured outputs, and multimodal inputs.