Firefox Voice - special opportunity for contributors

At Mozilla we are always looking at ways that we can use our voice to help people on the web. Whether it be through campaigning on important issues, or through helping protect users privacy on the web, Mozilla has used its voice to help people all over the world. Many of these projects are things that we, as contributors, are already involved in and are part of. But what if you just wanted to use to voice to close a tab in Firefox, or search for music, or have a page read to you? What if instead of typing into a search box, you could just ask Firefox to “search for sushi in Auckland”, or “find pictures of red pandas"?

The Emerging Technology team at Mozilla has been working on making this happen and need our help. They have created Firefox Voice to bring voice control to Firefox desktop. You click an icon on the toolbar (it only records when you want it to), speak a command, and Firefox does it. Behind the scenes, the audio is being processed by Google, but please appreciate that Google is not retaining your voice data; we hope this processing will be handled using Mozilla technology in the future (See the Mozilla Deep Speech project). The telemetry data that is retained and managed in line with Mozilla’s privacy policy and is not being stored with your personally identifiable information.

By being part of the advance testing group, you will be sent a link to download the add-on and may be sent occasional questions about the experience using it. You can download and install Firefox Voice from here.

As with any pre-Release software, as early testers we have the opportunity to help the team working on it by finding and filing bugs and making feature requests. You can do this here or directly into GitHub here.

I do appreciate that you may be a little concerned that this is only available in English - I too believe that the web should be available for all. Although the language choice is limited at present, the hope is to make use of the amazing work done by contributors to Common Voice. If you want to use your voice to help with this, you can do so here.

I hope that in reading this, you too are as excited about this development for Firefox as I am and the possibilities it may bring to help more people use the web.


Thanks for sharing the opportunity to use Firefox Voice!

Will it be help to add some sentences for voice control to Common Voice for people to recording them?

Just come up with a pre-defined sentences list, we can localized it, and than put them into Common Voice.


Hi @irvin

I think that is a great question, but I am not quite sure how we would go about this as there are so many permutations of what can be used in Firefox Voice.

As I mentioned, the plan is to make use of what Common Voice is achieving, but maybe more could be done. Looping in @nukeador - could a separate data set be put together for Firefox Voice (assuming the Firefox Voice project team could use it)?

1 Like


I think the end goal is use DeepSpeech models for Firefox Voice, but at the time of the release they were not polished enough for this.

@reuben might have more information

1 Like


Just a quick note to surface this topic now that it is outside of NDA (thank you project team!) and to remind people about the accompanying Activate campaign at: