Amazon is bringing Alexa+ to Italy with an ambition that goes well beyond updating its voice assistant. The company is attempting to reposition Alexa as something closer to a domestic operating system: a conversational layer that not only responds to requests, but actively executes tasks across devices, services and everyday routines. After its debut in the United States last year, Alexa+ is now arriving in Italy in a phased “early access” rollout.
Only Port-In
5G Max speed 2 Gbps
5G Portin 599
150GB in 5G Data
€5.99 /30 Days
A staged rollout built around Echo devices and Prime
Access to Alexa+ will not be immediate or universal. Instead, Amazon is introducing it through an early access program that functions like an extended beta. New compatible Echo devices will receive instant activation, while existing users will need to register and wait for an invitation as access gradually expands.
The service will be free during the early access period. After that, Amazon plans to price Alexa+ at €22.99 per month, though it will be included at no additional cost for Amazon Prime subscribers — effectively bundling the assistant into an already expanding subscription ecosystem that includes fast delivery, streaming services and digital perks.
Compatibility is broad, but not universal. Alexa+ will run on many existing Echo devices, but not on earlier generations such as Echo Dot (1st gen), Echo (1st gen), Echo Plus (1st gen), Echo Show (1st and 2nd gen), and Echo Spot (1st gen), which will remain on the standard Alexa experience.
Amazon is instead focusing its AI ambitions on newer hardware, including Echo Show 11, Echo Show 8, Echo Dot Max and Echo Studio. These devices are equipped with the AZ3 neural processor, designed to improve local computation and reduce reliance on cloud processing.
Fire TV devices are also part of the rollout, including the Fire TV Stick 4K Select, Fire TV Cube (3rd gen), Fire TV Stick 4K Max (2nd gen) and Fire TV Stick 4K Plus (2nd gen). Fire tablets and older Alexa-integrated devices are not supported.
A more natural way of speaking to machines
The most immediate change with Alexa+ is not technical but linguistic. Users are no longer required to issue rigid commands or repeatedly invoke the wake word. Instead, the assistant is designed to handle continuous conversation, incomplete sentences, shifts in topic and implied meaning, all while preserving context.
The voice itself has also been redesigned to sound more natural and expressive, moving away from the segmented, command-driven interaction style that defined earlier generations of voice assistants.
From answering questions to taking actions
The real shift, however, is functional. Alexa+ is designed not just to provide information, but to act on it.
It can compare products, summarize reviews, track prices and complete purchases on Amazon with voice confirmation. When items run out, it can suggest replacements or prompt a reorder at the best available price. It can also help organize shopping lists based on household preferences and dietary constraints.
In the smart home, the interaction becomes more intuitive. Phrases like “it’s dark” can trigger lighting adjustments, while a request such as “clean here” can activate a robot vacuum in the room where the command was spoken. Complex routines can be created entirely by voice, combining multiple actions without opening an app.
Alexa+ is also designed to follow the user spatially. Music or tasks can move from room to room without specifying devices, simply by saying “here,” reflecting an effort to make location-aware computing feel implicit rather than explicit.
Memory, personalization and context
A central feature of Alexa+ is its ability to build contextual memory over time. It can learn preferences, routines and household dynamics — from music tastes to dietary habits — and adapt its responses accordingly.
It can also distinguish between different family members using voice and visual recognition, tailoring suggestions and reminders to individual users. The result is a system that aims to feel less like a shared device and more like a personalized assistant embedded in daily life.
A broader ecosystem of services and integrations
Entertainment and third-party integrations are also central to the redesign. Alexa+ works with Amazon Music, Spotify and Apple Music, shifting from explicit search to mood- and context-based recommendations.
On the services side, Amazon is opening the assistant to external partners. TheFork will be among the first integrations, enabling restaurant reservations through voice. Additional partnerships, including services like WeTaxi and Treatwell, are expected later. Each transaction processed through Alexa potentially represents a new revenue stream for Amazon, through commissions embedded in the platform.
A multi-model AI system powered by Bedrock
Under the hood, Alexa+ is not powered by a single large language model. Instead, it runs on Amazon Bedrock, the company’s cloud platform for orchestrating multiple AI models through a unified interface.
These include Amazon’s own Nova models and systems developed by Anthropic, such as Claude. A routing layer dynamically selects which model handles each request, balancing speed, cost and complexity. Simple tasks, such as turning on lights, are handled by lighter systems, while more complex reasoning is delegated to more capable models.
Local processing and speech understanding
The system still begins at the edge. A low-power neural network on the device detects the wake word “Alexa,” after which audio is processed using beamforming and echo cancellation before being sent to the cloud.
Beamforming uses multiple microphones to isolate the user’s voice from background noise, improving the quality of speech recognition and intent detection. The audio is then processed through ASR (automatic speech recognition) and NLU (natural language understanding) systems that interpret meaning rather than just words.
An “expert system” for real-world tasks
One of the most ambitious elements of Alexa+ is what Amazon calls its “expert system.” Rather than treating requests as single actions, the assistant breaks them down into sequences of subtasks and executes them across different services.
A request to book a restaurant, for example, might involve checking availability, selecting options based on preferences, completing the reservation and sending confirmations to other people — all within a single conversational flow.
For developers, Amazon has introduced multiple SDKs to integrate services into this ecosystem, including API-based actions, web-based automation and a multi-agent framework that allows Alexa+ to interact with other AI systems. Alongside this is Nova Act, a visual reasoning model capable of navigating websites and completing tasks even without dedicated APIs.
Italy as a linguistic stress test
Bringing Alexa+ to Italy required more than translation. Much of the adaptation was handled by Amazon’s R&D center in Turin, which has worked on language modeling since the earliest versions of Alexa.
The goal is not only to understand standard Italian, but also regional variations, idioms and context-dependent meanings. Words like “salute,” for example, can shift meaning depending on whether they are used as a toast, a response to a sneeze or a formal greeting — distinctions that matter in spoken interaction.
Privacy and control
Amazon maintains its established privacy model. The assistant activates only after the wake word, and users can review, delete and manage recordings through a centralized dashboard available in the app and on the web.
Users are also able to control retention periods and review exactly what has been stored or shared. The system is built on AWS infrastructure, with privacy controls positioned as a core part of the product experience.
Beyond the home
Alexa+ is not confined to Echo devices. It will extend to Fire TV, mobile apps and eventually web browsers, where it will function more like a full conversational AI environment capable of generating content, analyzing data and managing tasks across contexts.
One of its defining features is continuity: conversations can begin on a kitchen device, continue on a smartphone while shopping, and resume later without losing context.
A shift in ambition
With Alexa+, Amazon is attempting something larger than an upgrade. It is trying to redefine what a voice assistant is for — moving from reactive tool to proactive system.
Whether that vision holds will depend less on model architecture than on everyday reliability: whether the system can consistently execute tasks correctly, integrate smoothly with third-party services and remain invisible enough to feel indispensable rather than intrusive.