Amazon Opens Smart Home AI Toolkit Preview to Expand Alexa Plus

Amazon is opening a new chapter for its voice assistant by shifting from static commands to flexible, agentic orchestration.

Alongside details shared at the 2026 Alexa+ Partner Summit in Seattle, Amazon announced a major expansion of its AI-driven smart home platform. The company introduced a developer preview for its new Smart Home AI Toolkit, designed to help device makers build deep, conversational integrations that move beyond basic on and off toggles.

Alexa+, which launched to all U.S. customers in February 2026 for $19.99 per month (free for Prime members, with a usage-limited, text-only free tier available to non-Prime users via Alexa.com and the Alexa app), is now adding more advanced capabilities for developers. By leveraging advanced large language models (LLMs) and adopting open communication standards, Amazon is attempting to transform Alexa+ into the central orchestration engine for connected homes.

The Shift to Agentic Smart Home Control

For years, smart home assistants have operated on a rigid, single-command architecture. If a user wanted to set up their home for a movie, they had to create a manual routine, explicitly mapping out which lights to dim, which plugs to activate, and what volume to set.

Alexa+ aims to reduce this manual setup by introducing agentic capabilities that understand context and manage multi-step reasoning. Instead of executing pre-programmed commands, the assistant can now interpret natural-language requests, evaluate constraints, and coordinate across devices.

Amazon showcased this capability with two smart home examples:

Intelligent Appliance Management: A user can tell Alexa+, "my kid's soccer jersey could use a deep clean, but the tag says cold wash only." Instead of the user navigating settings or choosing cycles manually, Alexa+ can automatically select the "cold-water enzyme activation cycle with extra rinse and low spin" from the appliance's thirty available cycles.
Proactive Automation Suggestions: Alexa+ can learn customer habits and proactively suggest personalized automations without requiring manual configuration. For instance, the assistant can infer user preferences to suggest brewing coffee automatically the moment a morning alarm goes off.

This approach allows the assistant to map conversational outcomes directly to specific device capabilities, making interactions feel more intuitive and less reliant on memorized syntax.

Adopting the Model Context Protocol (MCP)

One of the most significant technical announcements is Amazon's adoption of the Model Context Protocol (MCP). The standard was originally developed and open-sourced by Amazon's partner, AI laboratory Anthropic, in November 2024. Today, MCP is governed under the Agentic AI Foundation, a directed fund of the Linux Foundation, with the governing board chaired by AWS executive David Nalley. Anthropic states that MCP's technical governance model remains unchanged, with individual maintainers continuing to drive decisions.

Alexa+ is late among major AI assistants in opening to third-party MCP servers, behind ChatGPT, Gemini, and Copilot. The shift comes roughly a year after AWS itself began shipping MCP infrastructure for developers. Adopting this open standard makes it easier for diverse software and hardware providers to interface with conversational AI models. On Alexa+, this architecture routes requests between Amazon's own Nova models and Anthropic's Claude models via Amazon Bedrock.

By integrating MCP support, Amazon is making it easier for device manufacturers to connect their products to Alexa+. The integration streamlines the development pipeline through several core mechanisms:

Standardized Interfaces: Developers can bring their own MCP servers or APIs to the platform. Alexa+ for Builders can inspect the server, propose an integration path, and deliver a simulator-ready package, allowing manufacturers to expose advanced device capabilities without writing custom routing layers.
Broad Ecosystem Integration: Service providers can connect their platforms to Alexa+ using MCP. Priceline, Lyft, Headspace, and Virgin Atlantic will be among the first brands to start building experiences using MCP later this year, alongside other partners such as Canva, Cengage, Crystal Dynamics, and Viator, as well as ground-transportation providers Brightline, FlixBus, and Greyhound, which are integrating via Distribusion.

14 Launch Partners Leading the Integration

To support this rollout, Amazon is collaborating with 14 initial partner brands. These manufacturers are already integrating their hardware with the new Smart Home AI Toolkit, spanning appliances, security, lighting, and home infrastructure:

Brand	Category	Focus / Role
Bosch	Appliances	Home appliances and smart kitchen systems
Whirlpool	Appliances	Smart laundry and cleaning appliances
iRobot	Cleaning	Robotic vacuums and home cleaning
Ecovacs	Cleaning	Robotic cleaning solutions
Yale Home	Security	Smart locks and home access
Eufy (under Anker Innovations)	Security	Smart home security cameras and devices
TP-Link Systems Inc.'s Tapo	Smart Ecosystem	Smart plugs, lights, and sensors
Govee	Lighting	Smart ambient lighting
Twinkly	Lighting	Decorative smart lighting
Delta	Plumbing	Smart faucets and shower fixtures
Moen Flo	Water Management	Leak detection and water monitoring
PetSafe	Pet Care	Smart pet doors and automated feeders
Pila Energy	Energy	Home energy management
SONOFF	Infrastructure	Smart switches and relays

This group of partners provides Amazon with a broad hardware baseline to test how Alexa+ coordinates actions across lighting, security, plumbing, and home utilities.

Seamless Commerce with Amazon Wallet

To support the agentic capabilities of Alexa+, Amazon is integrating its payment infrastructure directly into the conversational flow. The inclusion of Amazon Wallet allows the assistant to handle transaction-based tasks securely.

If a user wants to book travel or hire a task service, Alexa+ can coordinate the request and complete the transaction using the payment method saved in their Amazon account. Amazon plans to support this feature via initial integrations with partner services like Priceline, Cengage, TaskRabbit, Atom Tickets, and Fandango.

According to Amazon, users are engaging with services integrated with Alexa+ six times more from initial discovery to final booking.

Early Engagement Metrics

According to self-reported figures released by Amazon, the transition to Alexa+ has driven significant increases in user activity. In terms of general service usage, Amazon claims that Alexa+ users are engaging with the assistant two to three times more, spanning a broader range of topics. On the smart home side, the company reports a nearly 50 percent increase in smart home engagement during the first month after upgrading, alongside users setting up twice as many automated Routines.

Current Availability and Editorial Outlook

The integration paths for the new Smart Home AI Toolkit are currently available in Preview, while the broader Alexa+ for Builders program is open to select partners working directly with Amazon's team. Amazon has not specified regional or language support for the toolkit preview. Meanwhile, the base Alexa+ service continues its international expansion. As of late July 2026, the service has reached full availability in Canada (exiting Early Access on June 11) and remains in Early Access in markets including Brazil, France, Italy, and Spain.

While these updates represent a significant push toward a more agentic assistant, history suggests that smart home enthusiasts should maintain a degree of caution. Amazon has a track record of announcing developer initiatives and previews well ahead of consumer availability. Because these features are still labeled as Preview, a wide public rollout may take time. Furthermore, the success of this ecosystem relies heavily on third-party developers successfully deploying and maintaining their own MCP servers.