Amazon has followed through on its promise of upgrading its artificial intelligence (AI) video model, Nova Reel, to make it capable of generating videos of up to two minutes in length as opposed to just six seconds, with which it was introduced back in December last year. Amazon Web Services (AWS), in a blog post, claimed that the latest Nova Reel, Nova Reel 1.1, can generate “multi-shot” videos with “consistent style” across shots.
Users will be able to provide a prompt of up to 4,000 characters to generate a video of up to two minutes, which will be composed of six-second shots, noted AWS in a blog post. On top of that, Nova Reel 1.1 brings a new feature called “Multishot Manual”, which allows users to guide video generation more precisely using both an image and a text prompt.
As per the AWS blog, when provided with an image at 1280 x 720 resolution and a prompt of up to 512 characters, the model can produce videos consisting of up to 20 individual shots, offering enhanced control over shot composition.
Amazon Nova Reel: Availability
Nova Reel is currently accessible exclusively via AWS platforms and services, such as Amazon Bedrock, the company’s AI development suite.
Also Read
Access to the tool is limited, targeted at enterprises and businesses, and requires users to submit a request.
Other text-to-video AI models
Other companies, including OpenAI, Google, and Meta, have also introduced text-to-video AI models in the past, namely- Sora, Veo, and Movie Gen, respectively.
The prime differentiating factor between these models and Amazon AWS’ Nova Reel is that these are accessible to consumers. However, Nova Reel is made available exclusively via AWS platforms and services, which makes it targeted at businesses and enterprises only.

)