Amazon Nova foundation models deliver frontier intelligence and industry-leading price-performance, enabling you to build generative AI applications that are safe, reliable and cost-effective.
This repository contains code samples demonstrating how to use the Amazon Nova API, an OpenAI-compatible interface for Amazon Nova's latest models.
⚠️ Please note: The Amazon Nova API is not currently intended for production usage. See our API Documentation for rate limits and quotas. For higher rate limits and quotas, explore Amazon Nova models on AWS.
To get started, ensure you have an Amazon Nova API Key. You can get one at nova.amazon.com/dev.
If you are building on AWS, please see AWS Samples for Amazon Nova.
- Python 3.x
- Jupyter Notebook (for running .ipynb files)
- Nova API Key from nova.amazon.com/dev
- Clone the repository:
git clone https://github.com/amazon-nova-api/getting-started-with-nova-api.git
cd getting-started-with-nova-api- Create a virtual environment (recommended):
python -m venv .venv
source .venv/bin/activate # On Windows: .venv\Scripts\activate- Install required dependencies:
For basic notebooks (00-04):
pip install -r requirements.txtFor Nova Agent examples (optional):
pip install -r nova_agents/requirements.txt # Adds Nova Act for computer useFor LangChain examples (optional):
pip install -r langchain/requirements.txt- Configure your API key:
cp .env.example .env- Edit
.envand replaceyour_api_key_herewith your actual Amazon Nova API Key
⚠️ Important: Never commit your.envfile to version control
- Getting Started: Complete basic set up, make your first calls, and learn Amazon Novas new features
- Multi Turn: Go from single calls, to chaining promp turns to complete a conversation with Amazon Nova
- Multi Modal Understanding: Get started with Amazon Nova's multi modal understanding capabilities, like image, and video understanding
- Tool Use with Nova: Learn the basics of tool use with Amazon Nova
- Nova Strands Agent: Get your first taste of building a simple agent with Strands SDK
- langchain/: Advanced examples of building Nova Agents with LangChain
- strands/: Advanced examples of building Nova Agents with LangChain
- nova_agents/: Agentic application examples combining Nova Acts computer use capabilities
This library is licensed under the MIT-0 License. See the LICENSE file.