Amazon Nova

Amazon Nova Foundation Models

Optimize your AI applications with powerful, cost-effective foundation models

Try Amazon Nova

Build with Amazon Nova

About Amazon Nova

Amazon Nova models deliver frontier intelligence and industry-leading price-performance. With the most comprehensive suite of customization capabilities for any proprietary model family, organizations can customize Nova to deliver responses that reflect their industry expertise. Using the appropriate model, you can get optimal intelligence, latency, and costs for your needs. Through Amazon Bedrock, organizations can seamlessly build and scale generative AI applications with Nova that are safe, reliable, and cost-effective.

Meet Nova 2 models

Models designed to power your diverse AI use cases

Nova accelerates AI innovation across modalities and capabilities, including text, image, video, speech, and agentic AI.

Nova 2 Lite

Nova 2 Lite is a fast, cost-effective reasoning model for most everyday tasks. It offers developer controls for thinking effort, built-in tools such as code-interpreter and web grounding, remote MCP tool support, and a 1M-token context window. You can use the model for everyday AI workloads such as customer service chatbots, document processing, and business automation. Nova 2 Lite is also the best model to use with Nova Forge to build your own frontier intelligence for the most complex tasks.

Nova 2 Pro (Preview)

Nova 2 Pro is our most intelligent model for highly complex workloads. It excels at tasks that demand the highest accuracy for advanced multi-step reasoning, long-range planning, and sophisticated agentic workflows. It can be customized with Nova Forge and is the best teacher for model distillation for faster and more cost-effective models. You can use the model for highly complex tasks such as multi-document analysis, video reasoning, and software migrations. Nova 2 Pro is in preview with early access available to all Nova Forge customers. Please reach out to your AWS account team for access.

Nova 2 Omni (Preview)

Nova 2 Omni is the industry’s first reasoning model that processes text, images, video, and speech inputs while natively generating both text and images. It features reasoning capabilities, enables natural language image editing and speech transcription. The model can be utilized for a wide range of tasks, including marketing content creation, customer support call transcription, video analysis and metadata generation, document understanding, and multilingual support documentation with visual aids. Nova 2 Omni is in preview with early access available to all Nova Forge customers. Please reach out to your AWS account team for access.

Nova 2 Sonic

Amazon Nova 2 Sonic is our speech-to-speech model for natural, real-time conversational AI. It unifies speech understanding and generation into a single model and delivers industry-leading conversational quality and pricing. It offers best-in-class streaming speech understanding, efficient dialog handling, instruction following, and tool invocation accuracy. The model now supports seven languages and new capabilities such as polyglot voices, cross-modal interaction for seamless switching between voice and text within a session, asynchronous tool use, and an expanded context window of up to 1M tokens. You can use Nova 2 Sonic for tasks such as customer support, interactive learning, and voice-enabled personal assistants.

Nova Multimodal Embedding

Nova Multimodal Embeddings is a state-of-the-art embedding model designed to power agentic RAG and semantic search. It is the first unified embedding model that supports text, documents, images, video, and audio through a single model. It maps diverse content types into a unified space with leading accuracy, helping to break down data silos. Organizations can unlock the value of their unstructured data by replacing multiple specialized models with a single unified solution to power seamless search and retrieval across content types.

See Nova in Action

Trellix Automate Threat Investigation with Nova 2

Sumo Logic uses Nova 2 to Accelerate SecOps

Siemens uses Nova 2 for Enterprise Search

I want to

learn more about Nova 2 Lite

Nova 2 is a powerful and efficient model for building AI agents with advanced reasoning and tool-use capabilities. Together with LangSmith’s observability, evaluation, and deployment workflows, developers can build, test, and ship reliable agents quickly. We’re looking forward to seeing how developers use Nova in more complex, real-world agentic systems.

Harrison Chase

Co-founder and CEO, Langchain

Webex by Cisco and AWS are building several generative AI capabilities leveraging Amazon Bedrock, LEX, Q index, and Quick Suite to enhance productivity and streamline workflows. We are also currently testing Amazon Nova 2 Sonic for AI Receptionist for Webex Calling, Cisco's automated front-desk agent. Nova's unified speech-to-speech architecture offers ease of integration with bidirectional streaming APIs and natural language prompts. As an early user of the model, we particularly like the the model's expressive voices, async task completion and language switching capabilities that create a more natural, human-like experience for our customers.

Amey Parandekar

Vice President of Product & GM, CISCO

At Securonix, our AI-driven cybersecurity platform relies on advanced contextual intelligence to detect and respond to evolving threats in real time. With Amazon Nova 2 Lite, we’ve achieved a new level of performance and efficiency, gaining 33% faster response times and seamless global deployment across the US, EU, and APAC regions. The model’s enhanced contextual understanding allows us to power our AgenticMesh security solution with deeper, more accurate threat analysis and automated remediation. What’s most exciting is that Nova 2 Lite now handles complex security content flawlessly, enabling innovations that weren’t possible before.

Venkat Kotla

CTO, Securonix

Alexa uses Amazon Nova Sonic for its new Alexa Greeting experience on Ring doorbell cameras —a new capability that allows Alexa to greet visitors, take messages, and provide instructions on where to deliver packages on your behalf.

Scott Durham

Director, Software Development, Alexa Domains - Smart Home

A unified multimodal model that can reason across text, images, video, and speech, and then generate high-quality visuals from natural language, fits perfectly with how we see the future of Mugen AI. Instead of stitching together multiple tools, we can now handle analysis and creation in one unified workflow. We are particularly excited about Nova 2 Omni’s ability to perform precise natural-language edits and handle complex creative tasks, which will help us streamline how we design, test, and scale marketing experiences. Add speech and video understanding on top, opens a new class of creative and analytical possibilities we are excited to experiment with.

Satoru Yamamoto

Chief AI Officer, Dentsu Digital

Nova 2 Lite provides a powerful new capability for building advanced AI systems with reasoning and tool-use. AWS Transform is a multi-agent, multi-user system that uses dozens of tools to modernize complex infrastructure systems and codebases. We breakdown each user request into a number of agent and tool calls that are routed to a mixture of models to balance cost, speed, and accuracy. Our evaluations show Nova 2 Lite improves tool calling efficiency for code modernization in AWS Transform by up to 60%. We are excited that AWS Transform will use Amazon Nova 2 Lite for code modernization workloads.

Asa Kalavade

VP of AWS Transform