Firefox Voice is a new experiment by Mozilla that has just been launched as a beta. The experiment brings voice controls to the Firefox web browser and is currently only available for desktop versions of Firefox.
Interested users may visit the Firefox Voice Campaign website to join the beta and install the Firefox Voice extension in the Firefox web browser.
The extension requires a truckload of permissions but that is understandable considering that you control different parts of the browser using it. You are asked if you want to allow Firefox Voice to collect voice transcripts and make them available to Mozilla for research purposes; you can allow or deny the request.
Note that Firefox Voice makes use of the Google Cloud Speech Service; any voice command is submitted to Google’s service. Mozilla notes that Google does not record these commands.
Firefox Voice is not the first attempt to bring voice controls to Firefox. Mozilla launched an experiment called Voice Fill back in 2017 to “speak to” search engines.
Firefox Voice detects microphones connected to the computer automatically. Since you control the browser using voice input, it is necessary that at least one microphone is properly connected to the device.
As far as options are concerned, there is a lot that you can use Firefox Voice already for even in this early stage of development:
- Search the web, e.g. “Search for Japanese restaurant in New York”, or “Look up recipes for lasagna”.
- Search specific websites, e.g. “Search my Gmail for tickets to event XYZ”, “Search CSS grid on MDN”. Only a selection of sites are supported including several Google services, Amazon, Facebook, Twitter, Reddit, Netflix, or Spotify.
- Go to website, e.g. “Go to Wikipedia”, or “Show me the 49ers schedule”.
- Ask questions, e.g. “who created Star Wars”, or “Who won the Grand Slam in 2019”.
- Play music or video, e.g. “Play Time in a Bottle on Spotify”, or “Play the Board Game Geek show”.
- Control media playback, e.g. “Play”, “Mute”, or “Next”.
- Read to me, e.g. “read this page”.
- Weather, e.g. “show me weather in Berlin, Germany”, or “What is the temperature in Bangkog?”.
- Translate, e.g. “Translate this webpage to German”, or “How do you say ‘phrase’ in Spanish?”.
- Traffic and Maps, e.g. “Find the nearest Thai restaurant on maps” or “How do i get to the nearest University”.
- Timer, e.g. “set a timer for 10 minutes”.
- Find tab, e.g. “Find calendar tab”.
- Browser controls, e.g. “Close tab”, “Open new tab”, or “Print”.
- Clipboard activity, e.g. “Copy title”, “Copy link”, or “Paste”.
- Giving commands nicknames, e.g. “Give that the name news” to assign “news” to the active site.
As you can see, Firefox Voice supports quite a good range of voice commands already. It is likely that the service will see even more commands in the future as development is still ongoing.
Firefox Voice is limited to controls in the English language currently. To use it, activate the Firefox Voice icon in the address bar and speak to control the browser using your voice.
Voice worked surprisingly well during an initial test. How well it works depends on a number of factors, including the microphone that is used, the command that is issued, and your pronunciation.
Additional information is available on the project’s GitHub webpage.
Firefox Voice is an interesting feature that works really well already. While I have little use for it, I can see its use for others. Obviously, having to click a button to run voice commands is not the most effective way of controlling the browser but since most users would not probably want that the service listens to everything, it is necessary to limit voice interaction.