Amazon’s next version of Alexa is being shaped around a difficult goal: turn a familiar free voice assistant into a more capable generative AI product that some users may be willing to pay for. According to Reuters, that effort is now expected to rely heavily on Anthropic’s Claude models after Amazon’s own models did not perform well enough in early versions.
Claude moves to the center of gen AI Alexa
Reuters reported that the previously announced generative AI version of Amazon’s Alexa voice assistant “will be powered primarily by Anthropic’s Claude artificial intelligence models.” The publication cited five anonymous people “with direct knowledge of the Alexa strategy.”
The shift matters because Amazon had already shown a generative AI Alexa in September 2023. At the time, the company described the assistant as more advanced, more conversational and able to handle multiple smart home tasks through simpler commands.
But Reuters reported that Amazon will no longer use its own large language models as the main engine for the new Alexa. One source told Reuters that early versions based on Amazon’s AI models “struggled for words, sometimes taking six or seven seconds to acknowledge a prompt and reply.” Reuters’ sources said Claude outperformed Amazon’s proprietary software, though the report did not specify which versions or features were compared.
Amazon did not deny that third-party models may be part of the assistant. In a statement to Reuters, the company said it uses different technologies for Alexa and that it begins with models built by Amazon while also using models from partners. The statement named Titan and future Amazon models as part of that mix.
The Anthropic connection raises business questions
Amazon has invested $4 billion in Anthropic, and UK regulators are currently investigating this. The source article says it is uncertain whether Amazon’s investment means Claude can be used for Alexa for free. Anthropic declined to comment on Reuters’ report.
That uncertainty is important because the new Alexa is also expected to come with a subscription fee. Alexa has reportedly lost Amazon tens of billions of dollars throughout the years, and the generative AI version is being treated as a major attempt to change the economics of the product.
The core issue is simple: a better Alexa may cost more to provide, but Alexa users have long associated the assistant with no direct monthly charge. The free version of Alexa is expected to remain available after the subscription version releases, which means Amazon may need to prove that the new features are meaningfully different from what users already have.
Pricing and launch timing remain unsettled
Earlier reports said the updated voice assistant would arrive in June, but Amazon still has not confirmed an official release date. The Washington Post reported on Monday that Amazon wants to launch the new Alexa in October, citing internal documents. Reuters’ sources said that timing could slip if the assistant does not meet certain unspecified internal benchmarks.
Pricing also appears unfinished. The Washington Post said gen AI Alexa could cost up to $10 per month, according to documents. That lines up with a June Reuters report that said the service would cost $5 to $10 per month. The Post also said Amazon would finalize pricing and naming in August.
Those details show the pressure around the product. Amazon needs the assistant to be capable enough to justify a subscription, but the company also has to choose a price that users will accept in a market where free alternatives exist.
What Amazon may offer subscribers
The Post’s report said the new Alexa will try to win subscribers with features such as AI-generated news summaries. A feature called Smart Briefing would reportedly provide summaries based on user preferences, including politics.
That is a sensitive area because the source article notes Alexa’s previous problems with reporting accurate election results. If Amazon uses generative AI Alexa for news summaries, the assistant’s reliability will matter as much as its convenience.
The Post also said the generative AI version would include “a chatbot aimed at children” and “conversational shopping tools.” Those reported features point to Amazon’s broader challenge: the company is not only trying to make Alexa answer more naturally, but also trying to make the assistant useful in more parts of daily life.
- Smart home control: Amazon previously promoted simpler commands for multiple smart home tasks.
- News summaries: Smart Briefing would reportedly tailor AI-generated summaries to user preferences.
- Shopping: The Post reported that conversational shopping tools are part of the plan.
- Children’s experience: The Post also reported “a chatbot aimed at children.”
The subscription hurdle
Some Amazon employees are questioning whether people will really pay for Alexa, Reuters noted. That concern is not hard to understand. A voice assistant that has been free has to clear a higher bar when it becomes a monthly product.
Amazon is also facing competition from other AI offerings, including free ones like ChatGPT. For gen AI Alexa, the question is not only whether the assistant can respond faster or sound more conversational. It is whether enough people will see it as worth paying for.
In June, Bank of America analysts estimated that Amazon could make $600 million to $1.2 billion in annual sales with gen AI Alexa, depending on final monthly pricing. That estimate assumed 10 percent of an estimated 100 million active Alexa users would upgrade. Amazon says it has sold 500 million Alexa-powered gadgets.
But those same analysts noted that free alternatives would challenge adoption. That makes the Claude report more than a technical detail. If Amazon is leaning on Anthropic’s model because it performs better, the decision may be central to whether the paid Alexa can feel different enough from the free assistant users already know.