Today, we’re announcing the launch of the Jamba 1.5 Model Family  — AI21 Labs’ new family of open models — in public preview on Vertex AI Model Garden. The model family includes two models designed for scaled enterprise applications:  

• Jamba 1.5 Mini: AI21’s most efficient and lightweight model, engineered for speed and efficiency in tasks including customer support, document summarization, and text generation.

• Jamba 1.5 Large: AI21’s most advanced and largest model that can handle advanced reasoning tasks — such as financial analysis — with exceptional speed and efficiency. 

Both models in the Jamba 1.5 Model Family offer a 256K context window, Mamba-Transformer architecture for efficient processing, and support advanced developer features like function calling, Retrieval-Augmented Generation (RAG) optimizations, and structured JSON output.

The combination of the 256K context window and efficient architecture allows these models to excel in handling key enterprise use cases such as summarizing and analyzing lengthy documents, powering RAG-based solutions, and a wide range of applications that demand both high-quality output and efficiency:

• Customer service: Help improve customer satisfaction and reduce costs through virtual assistants that handle inquiries across sectors like retail, healthcare, and financial services.

• Financial analysis: Summarize financial statements, extract key insights from market data, and generate comprehensive financial documents like loan term sheets to support quicker, more informed decisions.

• Content creation and summarization: Summarize large documents and generate relevant, high-quality text for content needs like product descriptions and FAQs.

These model additions on Google Cloud continue our commitment to an open and flexible AI ecosystem that helps you build solutions best suited to your needs. Google Cloud’s enterprise AI platform, Vertex AI, provides a curated collection of first-party, open source, and third-party models, many of which — including the new Jamba 1.5 Model Family — can be delivered as fully managed Model-as-a-Service (MaaS) offerings. With MaaS, you can choose the foundation model that fits your requirements, access it simply via an API, build with robust development tools, and deploy on our fully managed infrastructure — all with the simplicity of a single bill and hassle-free infrastructure. 

Experimenting and building with the Jamba 1.5 Model Family on Google Cloud

Google Cloud’s Vertex AI is a comprehensive AI platform for experimenting with, customizing, and deploying foundation models. AI21’s new models join over 150 models already available on Vertex AI Model Garden, further expanding your choice and flexibility to choose the best models for your needs and budget, and to keep pace with the continued rapid pace of innovation. 

By building with the Jamba 1.5 Model Family on Vertex AI, you can:

• Build with advanced development tools: Explore AI21’s models through simple API calls or evaluate the model with the Gen AI Evaluation Service in an intuitive environment.
• Optimize performance and costs with fully managed infrastructure: Scale your Jamba 1.5 Model Family applications from experimentation to production using Google Cloud's AI-optimized infrastructure. Benefit from pay-as-you-go pricing and auto-scaling to manage costs effectively while meeting enterprise-level performance demands.
• Deploy confidently with robust security, data privacy, and compliance: Protect the security and privacy of your data and AI applications with Google Cloud's comprehensive security features, data privacy controls, and compliance certifications.
• Craft intelligent agents: Create and orchestrate agents powered by the Jamba 1.5 Model Family using Vertex AI's comprehensive set of tools, including LangChain on Vertex AI.

Get started with the Jamba 1.5 Model Family on Google Cloud

1. Select the Jamba 1.5 Mini or Jamba 1.5 Large model tile in Vertex AI Model Garden. You can also find and easily procure Jamba 1.5 Mini and Jamba 1.5 Large on Google Cloud Marketplace and take advantage of the ability to draw down on your Google Cloud spend commitments.

2. Select “Enable” and follow the proceeding instructions.

3. Use our sample notebook to get started. You can also explore the Jamba 1.5 Model Family on Vertex AI documentation for further model details and code samples.

We're committed to providing developers with easy access to the most advanced AI capabilities. Our partnership with AI21 is a testament to Google Cloud’s commitment to provide you with world-class innovation in AI supported by an open and accessible AI ecosystem. We'll continue to work closely with our partners to keep our customers at the forefront of AI capabilities.