Amazon Nova foundation model
Amazon Nova Reel
Amazon Nova video generation model that accepts text and/or image input and outputs short videos through Amazon Bedrock.
Key points
- Generates short videos from text and/or image input.
- Supports video creation from prompts or source images.
- Is a creative Nova model, not a text-output understanding model.
- Is accessed through Amazon Bedrock.
When to use it
- Choose Reel when the required output is a short video.
- Use it for video creation from prompts or images.
- Use through Amazon Bedrock for managed Nova video generation.
Exam tips
- Reel is the Nova model for video generation.
- Do not confuse video input understanding in Lite, Pro, or Premier with video output generation in Reel.
- Canvas generates images; Reel generates videos; Sonic handles speech-to-speech.