Amazon’s new Nova AI models to take on OpenAI and Google with text, image, and video generation
Nova is expected to be 75% cheaper than its competitors and supports over 200 languages.
Amazon has been steadily weaving AI into its ecosystem throughout 2024. From debuting its AI shopping assistant Rufus in February to Maestro, a playlist generator launched in April, and most recently investing a whopping $4 billion dollars in Claude, the company’s ambitions to dominate the AI space have been clear.
Now, at its re:Invent conference held yesterday, Amazon Web Services (AWS) introduced Nova, a new family of generative AI models designed to process text, images, and videos.
The Nova lineup includes four text-generating models—Micro, Lite, Pro, and Premier—tailored for tasks ranging from basic text generation to complex reasoning. The first three are available now on AWS, while Premier is set to launch in early 2025. These models boast varying capabilities, with Lite and Pro supporting multimodal inputs like video and images. Meanwhile, Premier is described as a "teacher" model, enabling businesses to fine-tune custom models.
Emmanuel Oyedeji
Amazon has also unveiled two media-focused models: Nova Canvas, for creating and editing images, and Nova Reel, which generates six-second videos from prompts. AWS promises future updates to extend Reel’s capabilities to two-minute videos. These tools are integrated into Amazon Bedrock, a platform that allows customers to test and fine-tune AI models for their specific needs.
While AWS CEO Andy Jassy emphasized Nova’s speed, cost-effectiveness, and support for over 200 languages, questions linger about how Amazon’s AI tools compare to competitors like OpenAI’s GPT-4 and Google Gemini. Both rivals have long offered text and image generative capabilities with Gemini even able to generate videos and ChatGPT being able to search the web for you.
A video created by Nova Reel (Credit: AWS)
Perhaps the major difference would be in pricing. Nova models are said to be up to 75% cheaper than other leading models, but Amazon remains vague about the data used to train them. AWS further shared that it has implemented safeguards like watermarking and moderation tools to address potential misuse, but its lack of transparency around data sources could raise concerns for enterprises.
By expanding Nova’s features to include tools like Canvas and Reel, Amazon is positioning itself as a leader in creative AI applications. With future plans for a speech-to-speech model and an "any-to-any" multimodal system, Nova reflects Amazon’s vision of an AI-driven future.
For businesses using AWS, these advancements could reshape how AI is integrated into daily operations, offering new possibilities while raising questions about cost and competition.
Emmanuel Oyedeji
