Today, Amazon announced Amazon Nova, a new generation of foundation models (FMs) available exclusively through Amazon Bedrock. These models offer cutting-edge AI capabilities and industry-leading price-performance ratios. Here's a summary of the key features:
Understanding Models
1. Amazon Nova Micro
- Text-only model with lowest latency and very low cost
- 128K token context length
- Ideal for tasks like summarization, translation, and simple coding
2. Amazon Nova Lite
- Low-cost multimodal model for image, video, and text inputs
- Processes up to 300K tokens
- Suitable for real-time customer interactions and document analysis
3. Amazon Nova Pro
- Highly capable multimodal model with optimal accuracy, speed, and cost
- 300K token input context
- Excels in complex tasks and financial document analysis
4. Amazon Nova Premier (coming in early 2025)
- Most advanced multimodal model for complex reasoning
Creative Content Generation Models
1. Amazon Nova Canvas
- State-of-the-art image generation model
- Produces high-quality images with precise control over style and content
2. Amazon Nova Reel
- Video generation model for creating short videos from text prompts and images
Key advantages of Amazon Nova include:
- Excellence in Retrieval-Augmented Generation (RAG), function calling, and agentic applications
- Customization capabilities for enterprise-specific needs
- Built-in safety controls and watermarking for responsible AI use
These models are designed to lower costs and latency for various generative AI tasks, from document processing to creating marketing content and building AI assistants.
Amazon has announced a new set of artificial intelligence platforms, called foundation models, which include Nova for generating videos and Canvas for creating images from text prompts. The new services aim to help developers automate tasks and improve latency and cost. Amazon's CEO Andy Jassy emphasized the importance of responsible use of these tools, including watermarking to prevent the spread of harmful content. The move is seen as Amazon's biggest step in developing AI applications, as it competes with rivals like Adobe and Meta in the space.
LAS VEGAS (Reuters) -Amazon announced a new slate of artificial intelligence platforms, known as foundation models, at its annual AWS conference, allowing for text, image and video generation among other things.
The new offerings pit it against rivals like Adobe and Meta that are racing to serve customers who want to automate more of their services.
Amazon's CEO Andy Jassy announced the new "Nova" models on Tuesday. Developers, Jassy said, had a list of desires, prompting the new services.
"They want better latency. They want lower cost. They want the ability to do fine tuning," Jassy said.
The announcements at the conference in Las Vegas are Amazon's biggest step toward combating a reputation that it was caught flatfooted in developing AI applications as competitors sped ahead.
Video generation from a single image or text prompt has been particularly hot, with Adobe, Meta, OpenAI and Tik Tok parent ByteDance, among others, all announcing new such AI applications. Amazon on Tuesday said its Nova Reel software allows users to make six-second videos that can be useful for, say, displaying products on the Amazon website. Videos of up to two minutes will be available in the coming months, Amazon said.
Entertainment industry technologists are eager to get their hands on such tools to more efficiently enhance and expedite filmmaking capabilities. Still, others worry that such systems could infringe on copyrighted works.
Also on Tuesday, Amazon said it had developed Canvas for generating images from short text prompts. Jassy emphasized that Amazon would include watermarking to ensure the software is used responsibly to prevent harmful content from being spread.
Other offerings announced Tuesday are meant to speed the time it takes to process and analyze text.
https://www.aboutamazon.com/news/aws/amazon-nova-artificial-intelligence-bedrock-aws
Comments
Post a Comment