Alexa+

The AI industry is shifting focus from the smartest models to cost efficiency, prioritizing 'intelligence per dollar' as many models become sufficient for business tasks. Companies like Amazon are optimizing expenses by using less-powerful but cheaper AI for routine tasks while reserving pricier models for complex needs. Experts highlight factors like reliability, cost structure, and reusability to evaluate AI models based on value.

Amazon is reducing its reliance on Anthropic's Claude models for Alexa+ to lower AI costs, instead using in-house models and optimizing GPU efficiency. Internal forecasts showed Alexa+'s AWS cloud costs could nearly triple in 2026, prompting cost-saving measures like rerouting requests to cheaper models and improving computing capacity utilization.

Amazon is developing an Alexa AI project called Moonraker to handle complex, multi-step tasks, positioning Alexa in the AI agent race. However, internal documents reveal the project's high costs, with projected GPU expenses exceeding $100 million by 2026, and Amazon faces challenges including delayed rollouts and technical issues like inconsistent responses.

Amazon's Alexa+ can generate custom AI podcasts on demand. The company is expanding its assistant into a personalized AI content platform. This new feature allows for the creation of custom podcast episodes.

Amazon is combining its Rufus AI shopping assistant with its Alexa+ platform. This move comes after CEO Andy Jassy recently praised Rufus. The integration aims to improve service.

Amazon has launched a new AI shopping assistant called Alexa for Shopping in its search bar, replacing Rufus assistant. This personalized assistant is powered by Alexa+. The new feature aims to improve the shopping experience for Amazon users.

Alexa+ is now available on several Bose speakers, marking its debut on a non-Amazon device. This integration expands the reach of Amazon's upgraded smart assistant. The move allows Bose speaker users to access Alexa+'s features and capabilities.