The ecommerce giant announced the second generation of its Nova AI model at Re:Invent, a company conference held in Las Vegas. These models are nowhere near as popular as those offered by rivals like OpenAI and Google, but Amazon’s plans to make them highly customizable could see them gain traction among its cloud users.
Amazon detailed two improved larger language models, the Nova Lite and Nova Pro; a new realtime voice model called Nova Sonic; and a more experimental model called Nova Omni that presents a simulated type of reasoning using images, audio and video as well as text. The new models are being made available to a limited number of customers today.
More importantly, given the importance of its cloud business, Amazon is also releasing a tool called Nova Forge that will let customers create special Frontier models by adding their own training data to raw versions of the Nova 2 Lite and Pro models.
It’s already possible to fine-tune off-the-shelf AI models like Google’s Gemini and OpenAI’s GPT. But Amazon’s approach lets customers add data at different stages of model training, including the process of building a base model, a step known as custom pre-training that is typically reserved for larger AI labs.
“Everyone is looking for a frontier model that is an expert in its field,” Rohit Prasad, who leads Amazon’s AI efforts, told WIRED ahead of today’s announcements. Prasad says Amazon developed the technologies behind Nova Forge to empower internal teams, including developing Alexa and AI agents to build custom models. “It’s basically a new open training paradigm,” he says.
One customer that has already tested the approach is Reddit, which used Nova Forge to build a custom model to identify content that breaks the platform’s rules.
Fine-tuning traditional models won’t work, says Chris Slowey, Reddit’s chief technology officer, because most models are designed to avoid offensive or violent content entirely, meaning they’ll refuse to analyze certain content. Slowe says custom pre-training combined with traditional fine-tuning has created a Frontier model that is expert at understanding and using Reddit.
“Other LLMs understand Reddit as a concept, and how Reddit works, but they’re not confused,” says Slowe. “We actually built a Reddit expert model.”
<a href