Ministral-3-3B-Instruct-2512 and Ministral-3-8B-Instruct-2512 models now available on Amazon SageMaker JumpStart

Mistral AI's Ministral-3-3B-Instruct-2512 and Ministral-3-8B-Instruct-2512 models are now available on Amazon SageMaker JumpStart. Designed for edge deployment and resource-constrained environments, they enable AWS customers to deploy high-performance, scalable AI solutions on AWS infrastructure. The Ministral-3-3B-Instruct-2512 model, with a 3.4B language model and a 0.4B vision encoder, fits within 8GB VRAM (FP8) and supports a 256K token context window. It provides vision analysis, multilingual instruction understanding (English, French, Spanish, German, Chinese, Japanese, Korean, Arabic, etc.), strict adherence to system prompts, and native function calling with structured JSON output under the Apache 2.0 license. The Ministral-3-8B-Instruct-2512 model, comprising an 8.4B language model and a 0.4B vision encoder, fits within 12GB VRAM (FP8) and uses an interleaved sliding window attention pattern for fast, memory-efficient inference. While offering the same vision, multilingual, agent, and function calling capabilities as the 3B model, it delivers more powerful inference and generation performance. Using SageMaker JumpStart, these models can be deployed in just a few clicks to address specific AI use cases. You can access the JumpStart model catalog via the SageMaker console or deploy models to your AWS account using the SageMaker Python SDK. For details on deploying and using foundation models with SageMaker JumpStart, see the Amazon SageMaker JumpStart documentation.

Why it matters

Amazon SageMaker JumpStart is a platform that allows AWS customers to easily deploy and use foundation models. These new models are particularly useful for developers and data scientists deploying AI applications in edge computing and resource-constrained environments.