What the world says, in your language.
Tolkr™ interprets live audio for your staff as it plays, and turns recorded programmes into finished dubs, subtitles and translated transcripts. It runs on your organisation's own dedicated systems, with no cloud in between.
The original in its own script, and the translation under it, while the speaker is still talking.
Your people can read along, or listen to a natural voice.
Keep the original under the voice, or turn it away and hear only the translation.
Recorded from Tolkr™ Studio. A press statement, 
French to 
English.
Local AI
The AI that listens, translates and speaks runs on your organisation's own systems. That one choice gives you three things a cloud service cannot.
No cloud service charges you per minute, per word or per request.
During use, no connection leaves the PC.
Captions and a natural voice follow the speaker as they talk.
Languages
Every language Tolkr™ offers works both ways, into English and out of it.
Follow audio as it plays, turn a recording into a finished dub, or bring Tolkr™ into a call.
Inside the video meetings you already hold.
Your side needs a Windows PC, headphones and a small audio driver. The others join as they always do.
What you deliver
Every recording and every live session gives your staff the translated words as well as the voice, in the formats editors, archives and broadcasters work with.
Translated text
Dubbed media
Broadcast specification
What changes
Most foreign language audio is never interpreted, because nobody can book a person for all of it. This is the work Tolkr™ is for.
01 / 05
Before
A press conference starts abroad.
02 / 05
Before
The archive was never worth a studio dub.
03 / 05
Before
A recording may not leave the building.
04 / 05
Before
The training video exists in one language.
05 / 05
Before
An interview arrives in a language nobody reads.
Choose your sector and see Tolkr™ at work in it.
Programmes that were never dubbed go out in sync, with tracks and subtitles for post.
Follow channels and press conferences live, then quote from the transcript.
Statements and calls translated on your own workstations.
Each sentence beside its original, with its time and speaker.
Captions on the room's screen, and course videos dubbed with subtitles.
Training videos dubbed for each branch, and meetings in two languages.
Staff and partners hear each other in their own language.
Or start from one jobPress conferencesLive channelsNewsroom monitoringArchive backlogTraining videosPodcastsAll 22 jobs
This is a real sentence from the recording at the top of this page, followed from the first sound to the finished caption.
Takes the sound from the PC, an app, a microphone or a file.
La première phase coûtera 40 millions d'euros.
Recognises the speech and writes it in the original language.
Phase one will cost 40 million euros.
Translates each sentence in the context of the conversation.
La première phase coûtera 40 millions d'euros.
Phase one will cost 40 million euros.
Speaks the translation and shows both lines as captions.
Want each step in depth? Read How it works.
What a broadcaster, a ministry or a law firm asks before anything else.
Arabic, 
Hebrew and 
Persian done properlyCaptions in the right script and direction, with long lines broken where the script allows.
See the language listRecognition, translation and the voice run on each of your workstations. No account is needed to translate, and sessions are ordinary files in folders you control.
During use, it makes no connection outside the PC.
We install Tolkr™ with you on the workstations that need it, ready for the people who use them.
How a rollout goesEach sentence starts with its original, and Tolkr™ Studio reports the result before the file leaves your desk.
Not answered here? Ask the people who build it
Yes. Captions and a natural voice follow the speaker as they talk. In a dub each sentence is placed at its original.
Windows 11. A graphics card makes it faster, and it also runs on the processor. Tolkr™ sets itself up for each workstation.
Captions in 130 languages, 70 understood as they are spoken, and a natural voice in 36. See the language list.
Yes. File mode dubs long recordings, continues after an interruption and shows the time left.
The video with the dubbed and the original sound tracks, the sound files, subtitles, and a report of every sentence, as you choose.
Yes. Live sessions are kept as transcripts in both languages, and from a file you can choose subtitles and the sentence list alone, in both languages.
Yes. Each workstation runs Tolkr™ by itself, with no shared server, and we install it with you on the workstations that will use it.
Yes. Join the meeting on your PC: you hear the others in your language, and they hear you in theirs. Everyone can also read the translation as captions. If you turn those captions on, the translated text goes to the meeting service.
Yes. Bring a short clip or a channel you need to follow, and we run Tolkr™ on it with you.
Bring a recording or a channel you need to follow. We run Tolkr™ on it with you.
Request a demo