Announcing Mistral AI’s Mistral Large 24.11 and Codestral 25.01 models on Vertex AI

Editor’s note: This announcement was originally published on November 21, 2024, announcing the general availability of Mistral Large 24.11 on Vertex AI. This announcement was updated on January 13, 2025, to reflect the availability of Codestral 25.01, Mistral AI’s newest model designed for code generation tasks. 

In July, we announced the availability of Mistral AI’s models on Vertex AI: Codestral for code generation tasks, Mistral Large 24.07 for high-complexity tasks, and the lightweight Mistral Nemo for reasoning tasks like creative writing. Now, we’re announcing the general availability of Mistral AI’s newest models on Vertex AI Model Garden: Mistral Large 24.11 and Codestral 25.01.

Mistral Large 24.11 is an advanced dense large language model (LLM) of 123B parameters with strong reasoning, knowledge and coding capabilities, and builds upon its predecessor with improved long context, function calling, and system prompt. The model is ideal for use cases that include complex agentic workflows with precise instruction following and JSON outputs, large context applications requiring strong adherence for retrieval-augmented generation (RAG), and code generation. 

Codestral 25.01 helps developers write and interact with code through a shared instruction and completion API endpoint. The model supports over 80 coding languages and specializes in low-latency, high-frequency tasks such as code completion, fill-in-the-middle (FIM), and testing. Codestral 25.01 features a more efficient architecture and an improved tokenizer than its predecessor, generating and completing code over 2.5 times faster.

You can access and deploy the new Mistral AI models on Vertex AI through our Model-as-a-Service (MaaS) or self-deployment offering today.

What can you do with the new Mistral AI models on Vertex AI?

By building with Mistral AI’s models on Vertex AI, you can:

• Select the best model for your use case: Choose from a range of Mistral AI models, including efficient options for low-latency needs and powerful models for complex tasks like agentic workflows. Vertex AI makes it easy to evaluate and select the optimal model.

• Experiment with confidence: Mistral AI models are available through our fully managed Model-as-a-Service offering on Vertex AI. You can explore Mistral AI models through simple API calls and comprehensive side-by-side evaluations within our intuitive environment.

• Manage models without overhead: Simplify how you deploy the new Mistral AI models at scale with fully managed infrastructure designed for AI workloads and the flexibility of pay-as-you-go pricing.

• Tune the models to your needs: In the coming weeks, you will be able to fine-tune Mistral AI’s models to create bespoke solutions, with your unique data and domain knowledge. 

• Craft intelligent agents: Create and orchestrate agents powered by Mistral AI models, using Vertex AI’s comprehensive set of tools, including LangChain on Vertex AI. Integrate Mistral AI models into your production-ready AI experiences with Genkit’s Vertex AI plugin.

• Build with enterprise-grade security and compliance: Leverage Google Cloud’s built-in security, privacy, and compliance measures. Enterprise controls, such as Vertex AI Model Garden’s new organization policy, provide the right access controls to make sure only approved models can be accessed.

Get started with Mistral AI models on Google Cloud

These additions continue Google Cloud’s commitment to open and flexible AI ecosystems that help you build solutions best suited to your needs. Our collaboration with Mistral AI is a testament to our open approach, within a unified and an enterprise-ready environment. Vertex AI provides a curated collection of first-party, open-source, and third-party models, many of which — including the new Mistral AI models — can be delivered as a fully managed Model-as-a-Service (MaaS) offering — providing you with the simplicity of a single bill and enterprise-grade security on our fully-managed infrastructure. 

To start building with Mistral AI’s newest models, visit Model Garden and select the Mistral Large (24.11) model tile  or the Codestral (25.01) model tile. The models are also available on Google Cloud Marketplace here: Mistral Large (24.11) and Codestral (25.01). 

You can check out our sample code and documentation to help you get started