frontiermultimodal

Amazon (Nova)

Amazon Nova — fast, cost-efficient multimodal models built natively for AWS Bedrock.

3
Models tracked
1994
Founded
Nova Pro
Flagship model
Seattle
HQ
Visit aws.amazon.com/bedrockDocumentation

About

About Amazon (Nova)

Amazon's Nova model family, launched in late 2024, is designed for production workloads on AWS Bedrock. Nova Pro, Lite, and Micro cover a range of cost and performance tiers. Nova Pro handles multimodal inputs — text, images, and video — with a 300K context window, while Micro provides the lowest-cost text-only option for classification and extraction. All Nova models are accessed exclusively through Amazon Bedrock.

aws.amazon.com/bedrock

Key highlights

  • Nova Lite is among the cheapest multimodal models at $0.06/M input tokens

  • 300K context window in Pro and Lite for large document processing

  • Native AWS Bedrock integration using IAM, VPCs, and CloudWatch

  • Video input support in Nova Pro for multimedia content workflows

  • Available in multiple AWS regions with data residency controls

Models

3 models tracked

Context windows, pricing, capabilities, and API identifiers — everything you need to choose the right model for your workload.

Flagship

Nova Pro

2024

300Kcontext
TextVisionCode

Amazon's flagship Nova model. Handles text, images, and video with a 300K context window. Designed for complex multimodal tasks where AWS ecosystem integration and cost efficiency matter.

  • Text + image + video input
  • 300K context
  • AWS-native IAM/VPC
  • Cost-efficient frontier
In: $0.80 / M tokens
Out: $3.20 / M tokens
amazon.nova-pro-v1:0

Nova Lite

2024

300Kcontext
TextVision

A very low-cost multimodal model for high-volume production. At $0.06/M input tokens, it's among the cheapest multimodal models available anywhere.

  • Extremely low cost ($0.06/M)
  • 300K context
  • Fast inference
  • Image understanding
In: $0.06 / M tokens
Out: $0.24 / M tokens
amazon.nova-lite-v1:0

Nova Micro

2024

128Kcontext
Text

The most cost-efficient model in the Nova family. Text-only, optimised for classification, summarisation, and structured data extraction at maximum throughput.

  • Cheapest text model ($0.035/M)
  • Ultra-fast
  • Classification & extraction
  • High-throughput
In: $0.035 / M tokens
Out: $0.14 / M tokens
amazon.nova-micro-v1:0

Use Cases

What Amazon (Nova) is best for

Each provider has specialisms shaped by their architecture decisions, training data, and API design. Here is where Amazon (Nova) performs best.

AWS-native AI workloads
Cost-sensitive high-volume pipelines
Document extraction at scale
Multimodal content analysis
Serverless AI with Lambda + Bedrock

Get started

Start building with Amazon (Nova)

Visit the official Amazon (Nova) platform to access API keys, documentation, and pricing details.

Visit aws.amazon.com/bedrockRead the docsBack to all providers