AWS unveils Nova AI models and enhanced customer control service

Amazon Web Services is introducing a suite of new, internally developed AI models and a dedicated service for enterprises to create customized versions.

During his Re:Invent keynote on Tuesday, AWS CEO Matt Garman unveiled Nova 2, a collection of four new AI models added to the existing Nova family.

The initial AWS Nova suite debuted at last year's annual conference, featuring four text-generation models and one image-generation model. This year, AWS is enhancing these models and launching a complementary service.

"The momentum has been truly remarkable," Garman stated during his keynote. "Nova is now used by tens of thousands of customers, from major marketing firms and tech leaders like Infosys, Blue Origin, and Robinhood, to innovative startups like NinjaTech AI. Today, we're making Nova even more powerful."

The four new models include Nova 2 Lite, a cost-efficient reasoning model. Reasoning AI models analyze information before responding, processing text, images, and video to generate text suitable for everyday tasks. Nova 2 Pro is a reasoning agent designed for highly complex tasks like coding, capable of handling text, images, video, and speech.

Nova 2 Sonic is a novel speech-to-speech model for conversational AI applications. Nova 2 Omni is a multimodal reasoning and generation model that can process images, text, video, and speech input to produce both text and images.

Alongside these model upgrades, AWS announced a new service called Nova Forge. According to CNBC, this service allows AWS cloud customers to build their own tailored, frontier versions of AWS Nova models, termed "Novellas," for an annual fee of $100,000. Enterprises can access pre-trained, partially trained, or post-trained models to further train on their proprietary data.

Join the Disrupt 2026 Waitlist

Secure your spot on the Disrupt 2026 waitlist for priority access when Early Bird tickets are released. Previous Disrupt stages have featured leaders from Google Cloud, Netflix, Microsoft, Box, Phia, a16z, ElevenLabs, Wayve, Hugging Face, Elad Gil, and Vinod Khosla. They are part of over 250 industry experts leading 200+ sessions designed to accelerate your growth and competitive advantage. You'll also have the chance to connect with hundreds of startups driving innovation across all sectors.

Join the Disrupt 2026 Waitlist

Secure your spot on the Disrupt 2026 waitlist for priority access when Early Bird tickets are released. Previous Disrupt stages have featured leaders from Google Cloud, Netflix, Microsoft, Box, Phia, a16z, ElevenLabs, Wayve, Hugging Face, Elad Gil, and Vinod Khosla. They are part of over 250 industry experts leading 200+ sessions designed to accelerate your growth and competitive advantage. You'll also have the chance to connect with hundreds of startups driving innovation across all sectors.

Garman explained that this approach aims to address challenges enterprises face when integrating their own data into pre-trained AI models.

"The more you customize models by adding data after initial training, the more they tend to lose some of the core reasoning capabilities they learned earlier," Garman said. "It's similar to humans learning a new language. When you're very young, it's relatively easy to pick up. Trying to learn a new language later in life is much more difficult. Model training exhibits a similar characteristic."

Early adopters of Nova Forge include companies like Reddit, Sony, and Booking.com.

Follow all of TechCrunch's coverage of the annual enterprise technology event here.