Developers building apps that deliver content spoken aloud by AI assistants have a new trick to implement with Amazon’s Alexa voice now. A new “long-form” speaking style (you can hear a sample of it here, compared to the neutral setting here) uses an updated text-to-speech model that’s optimized to read content like this blog post.
Converting content for people who prefer to listen to podcasts when they’re in the kitchen or driving can be done easily, with the new style that’s programmed to insert natural-sounding pauses between paragraphs or if it’s reading dialogue that belongs to different characters.
That ability is only available in US English for now, but Amazon also revealed that the Polly AWS AI service, which supports 29 languages, can use new conversation or news speaking styles. Polly also has 10 additional voices covering six new languages, like US Spanish, French Canadian and more.