Text-to-Speech for your Website – the 7 Best Solutions
Last Updated on
by Carl Robinson
In the ongoing effort to explore the reaches of voice space, this blog post will provide an overview of the top companies that offer voice tech services. These are AI-backed platforms that perform a wide range of services that include converting text to audio (text-to-speech or TTS), conversational and interactive voice needs, speech recognition, and everything in between.
What was once a technically challenging and costly undertaking available only to select few, is now an expanding industry whose affordability and timely delivery is helping voice technology usher in today’s age of efficiency. Below you’ll find the shortlist of top 7 said companies:
This Tel Aviv startup is at the forefront of the audio revolution. They provide an audio content solution, aiming to bring audio to the publishing sector, as well as to every content creator out there looking for a way to grow their audience, keep it engaged, and monetise it.
Once implemented, the audio player has an option to automatically create a customised set of smart speaker skills from selected content. Trinity Audio also has a content technology (or ConTech, as the company calls it) solution that constantly learns audience behaviour and provides audio content recommendations. TrinityAudio is also a supporter of independent media, and is a sponsor of the Voice Tech Podcast.
Founded in California, USA in late 2018, BingeWith’s converts online text content into rich audio through an embedded player, allowing users to instantly transform any blog and article into audio.
Alternatively, to create an mp3 file, you can paste your article link or the raw text into BingeWith’s website, and it will create an audio version that you can download.
BingeWith uses AI that matches natural speech patterns, taking cues from the text such as commas and semicolons, to generate natural sounding speech. There’s also a choice of voices from different regions and genders.
The embeddable player is attractive and its colour can be customised. It features nice big play button, 15s skip forward and back, and the total time of the audio clip. Listeners can also set the audio speed they feel most comfortable with. We currently use BingeWith on this site, so you can have a play with it at the top of this page!
Hailing from Minnesota, US, speak2web is a conversational voice assistant platform dedicated to providing services for businesses of all types and sizes, rather just than individual blogs. Operating under the paid subscription business model with a free trial available, the company offers a plugin that adds a voice assistant called Simon AI to any website, enabling visitors to engage in a conversation, just like they’re talking to a real person.
The set up takes up to an hour and users can choose between a male or female narrator.
Although it mainly focuses on WordPress, speak2web supports other website platforms and mobile apps. The company’s REST API solution easily integrates into any website or mobile app, third parties such as listings and reservation management systems, and even into unstructured data sources. The voice assistant gains access to all of the data on the website, allowing visitors to search and conduct purchases using voice input.
SpeechKit is an independent, open-source solution that adds text-to-speech audio to any site within seconds, as part of an all-in-one-toolkit for automated audio news publishing. With access to dozens of the latest voices from Google WaveNet and Amazon Polly, this platform provides natural language programming technology and audio analytics with detailed audio engagement reports.
Apart from audio editions of news articles and newsletters, SpeechKit also offers a custom Alexa Skill to deploy audio content through a WordPress plugin, RSS feeds or its audio API. The company recently launched a beta version of AdStudio (beta), its self-serve audio ad platform that gives users the ability to sell pre-roll audio sponsorships, which likely steers the service more towards established businesses than independent bloggers.
Play.ht is another TTS service that allows WordPress users to make their blog posts listenable as podcasts. It also allows Medium writers to add audio to their articles through a Chrome-only browser extension. By completing a simple setup procedure and clicking on the ‘Publish with Audio’ button, authors can instantly give voice to their written word in a wide range of voices and languages. They can either publish new articles with audio or subsequently add audio to already published articles on their Stories page.
The audio player itself is customisable to match your blog’s look and feel. Play.ht also provides a decent amount of audio analytics so users can check how many people are listening, sharing, and/or downloading their stories.
The SpeechSynthesis part allows webpages to read out their text content, and is supported by the majority of major browsers.
SpeechRecognition allows web pages to listen to visitors (through their microphone, if they give permission), recognise voice context from the audio input and to respond appropriately. SpeechRecognition is currently only fully supported by Chrome.
Together, these two parts facilitate communication between a webpage and a visitor. The TTS voices and the sections of text to be spoken by the system are fully customisable, but implementing the two-way interaction requires some coding skills.
An HTML5-based unlimited text-to-speech library, ResponsiveVoice also has a WordPress plugin designed to add voice features to the content management system across smartphones, tablets, and desktop computers. At the moment, it supports 51 languages and 168 voices.
When added to WordPress website, the player itself is non-invasive and looks like a simple button that says ‘Listen to Post’, ‘Listen to this’, or whatever you want it to say. The whole process only involves adding a shortcode anywhere in the page or post, and you’re good to go. You can set it to read a whole page or just portions of text. On top of that, you can set other parameters, such as voice pitch, rate, and volume.
Since 1999, ReadSpeaker have been one of the leading B2B TTS companies worldwide with over 10,000 customers in many different verticals. They develop branded custom voices for worldwide known brands, and have 90 high quality custom TTS voices in over 30 languages. Their products product portfolio covers Cloud, Production and On-Premise solutions for speech enabling IVRs, conversational systems, voice assistants, automotive, robotics, public announcement systems, learning platforms, websites and more.
They also created the first TTS application that speech enables websites. You can hear a demo of their product here – simply click the link and a page from this blog will load, then click the ‘Listen’ widget under the image to play the audio. It even highlights the text that it’s reading, right on the page!
Carl is the host of the Voice Tech Podcast. Since launching in April 2018, Carl has conducted over 40 in-depth interviews with voice industry experts, building one of the most well known media brands in the voice AI space.
Carl is the author of Voice Chops Tuesday, the number 1 voice technology newsletter that is enjoyed by over 1000 voice tech fans each week.
Carl is also a startup founder, product manager and data scientist, recently presenting a model for voice emotion conversion at ICASSP 2019.
To contact Carl, use the links in this box to send an email, tweet or message.