TLDRs:
- AWS launches Nova 2 models supporting text, speech, image, and video tasks.
- Nova Forge enables organizations to train AI models with proprietary data.
- Nova Act allows browser automation with AI agents achieving high reliability.
- Nova 2 Lite outperforms key competitors, though cost and latency remain unclear.
Amazon has unveiled its latest suite of AI models under the Nova 2 brand, aiming to accelerate automation and AI adoption in enterprises.
The lineup includes Lite and Pro models designed for reasoning tasks, Sonic for speech-to-speech conversions, and Omni, which supports multimodal operations across text, images, video, and speech. This expansion positions AWS to offer businesses a versatile toolkit capable of handling a wide range of AI-driven workflows.
The new models are designed to integrate seamlessly into enterprise environments, allowing organizations to scale AI applications while maintaining security and privacy.
According to Amazon, early internal and customer tests show strong performance, with the Lite model surpassing or matching other leading AI models in several benchmark assessments.
Tailored AI with Nova Forge
To address the growing demand for customized AI solutions, AWS also launched Nova Forge, a service that allows businesses to train bespoke versions of Nova models.
Organizations can use their own datasets during multiple stages of training, including pre-training, supervised fine-tuning, and alignment for instruction-following and safety tuning.
Nova Forge enables companies to create AI models that closely reflect their domain-specific knowledge and operational needs. By leveraging proprietary data, enterprises can enhance model accuracy, ensure better contextual understanding, and maintain competitive advantages while keeping sensitive information secure on AWS infrastructure.
Automation Made Simple with Nova Act
Alongside Nova Forge, AWS introduced Nova Act, a tool that allows users to build AI agents capable of automating browser-based workflows. These agents can perform repetitive or rule-based tasks with reported reliability of up to 90% in early customer trials.
Nova Act supports human-in-the-loop (HITL) interventions, allowing users to escalate complex tasks that the AI cannot resolve automatically.
Additionally, agents provide detailed execution logs, session recordings, and observability dashboards, enabling organizations to monitor, debug, and optimize AI-driven processes. The platform integrates with Amazon S3, making it easy to store and manage workflow artifacts.
Performance and Competitive Landscape
While the Nova 2 Lite model has demonstrated strong performance in benchmark tests, matching or exceeding competitors like Claude Haiku 4.5 (Anthropic), GPT-5 Mini (OpenAI), and Gemini Flash 2.5 (Google), AWS has not disclosed latency or cost-of-ownership data.
The company highlights that Nova Premier generates fewer tokens per answer, but comparable metrics for Nova 2 Lite and other models remain unreported.
Enterprise users will need to evaluate model efficiency, cost, and contextual capabilities for their specific applications. Despite these unknowns, the combination of multimodal abilities, custom training via Nova Forge, and automation through Nova Act makes AWS’s new offerings a powerful toolkit for businesses seeking to adopt AI at scale.
Conclusion
AWS is making a strong push into enterprise AI with its Nova 2 models, emphasizing versatility, customization, and automation.
From multimodal processing with Omni to tailored training in Nova Forge and high-reliability automation via Nova Act, businesses now have multiple pathways to integrate AI into daily operations.
While some benchmarking and cost details remain unclear, the Nova platform reflects AWS’s commitment to expanding AI capabilities across sectors.
Get 3 Free Stock Ebooks
Discover top-performing stocks in AI, Crypto, and Technology with expert analysis.
- Top 10 AI Stocks - Leading AI companies
- Top 10 Crypto Stocks - Blockchain leaders
- Top 10 Tech Stocks - Tech giants
