What is Amazon Nova?
Amazon Nova foundation models deliver frontier intelligence and industry-leading price-performance. Organizations can build and scale generative AI applications with Amazon Nova that are safe, reliable, and cost-effective. Amazon Nova accelerates AI innovation across modalities and use cases - including text, image, video, speech, API calling, and agentic AI. To start building with Amazon Nova, you must access the models through an API using Amazon Bedrock.
Amazon Nova delivers powerful foundation models across understanding, image, video, and speech capabilities.
Understanding models
Amazon Nova understanding models are designed to comprehend diverse inputs, including text, images, video, documents, and code. These models excel in powering interactive chat interfaces, Retrieval-Augmented Generation (RAG) systems, agentic applications, video analysis, and UI workflow automation solutions.
-
Amazon Nova Micro is a text-only model that delivers the lowest latency responses at very low cost.
-
Amazon Nova Lite is a very low cost multimodal model that is lightning fast for processing image, video, and text inputs.
-
Amazon Nova Pro is a highly capable multimodal model and our best combination of accuracy, speed, and cost for a wide range of tasks.
-
Amazon Nova Premier is our most capable multimodal model for complex tasks and our best teacher for distilling custom models for cost-effective applications.
Amazon Nova understanding models are available in US East (N. Virginia) and AWS GovCloud (US-West). Amazon Nova Micro, Lite, and Pro are also available in Asia Pacific (Sydney) and Europe (London). The models are accessible in the following regions via cross-region inference profiles: US West (N. California), US West (Oregon), Europe (Stockholm), Europe (Frankfurt), Europe (Ireland), Europe (Paris), Asia Pacific (Tokyo), Asia Pacific (Seoul), Asia Pacific (Mumbai), Asia Pacific (Singapore), and Asia Pacific (Sydney).
Creative models
Amazon Nova creative models transform text and image inputs into professional-grade images and videos. They are designed to deliver customizable visual content with control features, making creative content generation accessible and efficient for media, entertainment, retail, marketing, and advertising industries.
-
Amazon Nova Canvas is a high-quality image generation model with customization and control.
-
Amazon Nova Reel is a video generation model designed to democratize creative content generation.
Amazon Nova Canvas and Amazon Nova Reel are available in US East (N. Virginia), Europe (Ireland), and Asia Pacific (Tokyo).
Speech model
Amazon Nova Sonic is a foundation model for conversational speech understanding and generation. The model accepts speech as input and provides speech with text transcriptions as output. Amazon Nova Sonic offers a natural, human-like conversational AI experience with contextual richness. It is the first model to feature bidirectional streaming API capabilities, allowing for real-time, low-latency multi-turn conversations.
-
Amazon Nova Sonic delivers real-time, human-like voice conversations.
Amazon Nova Sonic is currently available in US East (N. Virginia), Europe (Stockholm), and Asia Pacific (Tokyo) and for English (US, UK) and Spanish.
For full model and region support information in Amazon Bedrock, see Supported foundation models in Amazon Bedrock
Overall model information
Amazon Nova Premier |
Amazon Nova Pro |
Amazon Nova Lite |
Amazon Nova Micro |
|
|---|---|---|---|---|
Model ID |
amazon.nova-premier-v1:0 |
amazon.nova-pro-v1:0 |
amazon.nova-lite-v1:0 |
amazon.nova-micro-v1:0 |
Inference Profile ID |
us.amazon.nova-premier-v1:0 |
us.amazon.nova-pro-v1:0 |
us.amazon.nova-lite-v1:0 |
us.amazon.nova-micro-v1:0 |
Input modalities |
Text, Image, Video |
Text, Image, Video |
Text, Image, Video |
Text |
Output Modalities |
Text |
Text |
Text |
Text |
Context Window |
1M |
300k |
300k |
128k |
Max Output Tokens |
10K |
10k |
10k |
10k |
Supported Languages |
200+1 |
200+1 |
200+1 |
200+1 |
Regions |
US East (N. Virginia)2 |
US East (N. Virginia)2, Asia Pacific (Tokyo)2, Asia Pacific (Sydney)2, Europe (London)2, AWS GovCloud (US-West) |
US East (N. Virginia)2, Asia Pacific (Tokyo)2, Asia Pacific (Sydney)2, Europe (London)2, AWS GovCloud (US-West) |
US East (N. Virginia)2, Asia Pacific (Tokyo)2, Asia Pacific (Sydney)2, Europe (London)2, AWS GovCloud (US-West) |
Document Support |
PDF, CSV, DOC, DOCX, XLS, XLSX, HTML, TXT, MD |
PDF, CSV, DOC, DOCX, XLS, XLSX, HTML, TXT, MD |
PDF, CSV, DOC, DOCX, XLS, XLSX, HTML, TXT, MD |
No |
Converse API |
Yes |
Yes |
Yes |
Yes |
InvokeAPI |
Yes |
Yes |
Yes |
Yes |
Streaming |
Yes |
Yes |
Yes |
Yes |
Batch Inference |
Yes |
Yes |
Yes |
Yes |
Fine Tuning |
No |
Yes |
Yes |
Yes |
Provisioned Throughput |
No |
Yes |
Yes |
Yes |
Bedrock Knowledge Bases |
Yes |
Yes |
Yes |
Yes |
Bedrock Agents |
Yes |
Yes |
Yes |
Yes |
Bedrock Guardrails |
Yes (text only) |
Yes (text only) |
Yes (text only) |
Yes |
Bedrock Evaluations |
Yes (text only) |
Yes (text only) |
Yes (text only) |
Yes |
Bedrock Prompt flows |
Yes |
Yes |
Yes |
Yes |
Bedrock Studio |
Yes |
Yes |
Yes |
Yes |
Bedrock Model Distillation |
Teacher to: Pro, Lite, and Micro |
Teacher to: Lite and Micro Student of: Premier |
Student of: Premier and Pro |
Student of: Premier and Pro |
1: Optimized for these 15 languages: English, German, Spanish, French, Italian, Japanese, Korean, Arabic, Simplified Chinese, Russian, Hindi, Portuguese, Dutch, Turkish, and Hebrew.
2: You can access this model in the US East (Ohio), US West (Oregon), Europe (Stockholm), Europe (Ireland), Europe (Frankfurt), Europe (Paris), Asia Pacific (Tokyo), Asia Pacific (Singapore), Asia Pacific (Sydney), Asia Pacific (Seoul), and Asia Pacific (Mumbai) regions through cross-region inference. Cross-region inference allows you to seamlessly manage unplanned traffic bursts by utilizing compute across different AWS Regions. With cross-region inference, you can distribute traffic across multiple AWS Regions. To learn more about cross-region inference, see Supported Regions and models for inference profiles and Improve resilience with cross-region inference in the Amazon Bedrock User Guide.
Amazon Nova Canvas |
Amazon Nova Reel |
|
|---|---|---|
Model ID |
amazon.nova-canvas-v1:0 |
amazon.nova-reel-v1:1 |
Input Modalities |
Text, Image |
Text, Image |
Output Modalities |
Image |
Video |
Max Prompt Length |
1024 characters |
|
Input Context Window |
512 characters |
|
Output Resolution (generation tasks) |
4.19 million pixels (that is, 2048x2048, 2816x1536) |
1280x720, 24 frames per second |
Max Output Resolution (editing tasks) |
Must meet all of the following:
|
|
Supported Input Types |
PNG, JPEG |
PNG, JPEG |
Supported Languages |
English |
English |
Regions |
US East (N. Virginia), Europe (Ireland), and Asia Pacific (Tokyo) |
US East (N. Virginia), Europe (Ireland), and Asia Pacific (Tokyo) |
Asynchronous Invoke Model API |
No |
Yes |
Invoke Model API |
Yes |
No |
Amazon Nova Sonic |
|
|---|---|
Model ID |
amazon.nova-sonic-v1:0 |
Input Modalities |
Speech |
Output Modalities |
Speech with transcription and text responses |
Context Window |
300K context |
Max Connection Duration |
8 minutes connection timeout, with max 20 concurrent connections per customer.1 |
Supported Languages |
English (US, UK) and Spanish |
Regions |
US East (N. Virginia), Europe (Stockholm), and Asia Pacific (Tokyo) |
Bidirectional Stream API Support |
Yes |
Bedrock Knowledge Bases |
Supported through tool use (function calling) |
1: By default, the connection limit is 8 minutes, however you can renew the connection and continue the conversation by providing the previous conversation's history.
