AI Models

Amazon Scales Back Nova AI Models, Shifts Focus to New Frontier Team

Amazon is reportedly scaling back most of its in-house Nova AI models, including the flagship Nova Premier and Omni models, the Reel video model, and the Canvas image generator. The models will remain online for existing customers but are no longer actively developed. The company is redirecting resources to a new Frontier Model Research group led by Pieter Abbeel, with a new foundation model expected at the re:Invent conference this fall.

Neura News

Neura News

Neura Market Editorial

July 28, 20262 min read
Amazon Scales Back Nova AI Models, Shifts Focus to New Frontier Team

Amazon is overhauling its artificial intelligence strategy and scaling back most of its in-house Nova AI models, according to a report from Business Insider that cites internal sources. The cuts affect the company's flagship models, including Nova Premier and Omni, as well as the Reel video generation model and the Canvas image generator. These models will remain available in a "keep the lights on" mode for existing customers, but Amazon is no longer actively developing them.

The pullback follows recent layoffs in Amazon's AGI division and the closure of the AGI Lab that the company founded in 2024 after acquiring the startup Adept. Amazon first unveiled its Nova family of AI models in December 2024, positioning them as a key part of its AI offerings.

Despite these cutbacks, Amazon is not exiting the AI race. The company is redirecting its resources to a new Frontier Model Research group led by Pieter Abbeel, a prominent AI researcher who joined Amazon through the acquisition of robotics startup Covariant. A new foundation model is expected to be unveiled at Amazon's re:Invent conference this fall.

A spokesperson for Amazon told Business Insider that AI models remain one of the company's most important projects. Amazon is also one of the largest backers of Anthropic and OpenAI, with investments that could potentially exceed $100 billion.

The #1 Newsletter in AI

Stay ahead of the AI curve

The most important updates, news, and content — delivered weekly.

No spam. Unsubscribe anytime.

The shift in strategy reflects a broader trend in the AI industry, where companies are increasingly focusing on a smaller number of high-impact models rather than maintaining a wide portfolio. Amazon's decision to concentrate its AI efforts on the Frontier Model Research group under Abbeel suggests a bet on fundamental research and long-term capabilities.

The Nova models, while being scaled back, still serve existing customers who rely on them. The company has not announced any immediate plans to discontinue the models entirely, but the lack of active development indicates that Amazon is prioritizing other areas of its AI business.

Amazon's investment in Anthropic and OpenAI underscores its commitment to the broader AI ecosystem, even as it refines its own internal efforts. The company's partnership with Anthropic has been particularly notable, with Amazon investing billions of dollars in the AI startup.

Related on Neura Market

More from Neura News

AI Models

42 Mathematicians Urge Royal Society to Warn Government and Media About AI Existential Risk

Forty-two mathematical fellows, including Fields Medal winners Martin Hairer, Peter Scholze, and Wendelin Werner, have signed an open letter urging the Royal Society to warn the UK government and media about existential risks from advanced AI. The letter follows recent breakthroughs in which leading models solved open research problems, including a Millennium Problem. None of the signatories are affiliated with AI companies. The group warns that AI labs' estimates of existential risk above ten percent must not be dismissed as hype, and that by the time the situation becomes obvious to the public, it may be too late to act.

Sep 18·2 min read
Developer

Steve Yegge Shuts Down Gas Town After Failing to Build Anything Else With It

Steve Yegge shut down Gas Town, his ultra-vibed coding agent orchestrator, after admitting he never built anything else with it despite heavy subscription spend. Databricks reported a 60% coding spend increase after rolling out GPT-6 Astra to 3,500 engineers, OpenAI published a misalignment disclosure framework with six case reports, and Xiaomi ran MiMo-V2.6 RL training in public with live telemetry.

Sep 18·21 min read