Skip to content
Jitsi.help
HostingServicesGuidesTroubleshootingComparePricing
Get helpDeploy in one click

Jitsi transcription setup service

Updated By the Jitsi Help engineering team

Short answer

We set up Jitsi transcription with Jigasi: live captions in meetings and a saved transcript per meeting. We use Vosk for free, fully private speech recognition on your own server, or Google Cloud Speech when you need more languages or accuracy. Transcripts are uploaded to your storage automatically.

What you get

  • Jigasi in transcriber mode with a working Prosody account
  • Vosk speech server on your infrastructure, or Google Speech
  • Live captions in meetings
  • A text transcript saved for every meeting
  • Automatic upload of transcripts to S3 or your storage
  • Right-sized server (16 GB RAM or more for Vosk)

Typical timeline

1 to 3 working days

Get a quote

Transcription turns every meeting into searchable text, and live captions make meetings accessible. On Jitsi it is also the component with the most undocumented pitfalls. We fixed them for our own platform, so we can fix them for you.

What we set up

  • Jigasi as a transcriber, not a SIP gateway, with its own authenticated account.
  • A speech engine: Vosk on your server for free and private, or Google Cloud Speech.
  • Captions and transcripts: live subtitles in meetings, a text file per meeting.
  • Storage: transcripts uploaded to your bucket alongside recordings.

The technical details are in our transcription guide.

On our platform

Transcription is a one-time $29 feature per server on our one-click platform, sized and configured automatically.

Frequently asked questions

Does audio leave our servers?

Not with Vosk. Speech recognition runs in a container next to Jitsi, so audio and transcripts never leave your infrastructure. With Google Speech, audio is sent to Google's API for recognition.

Which languages are supported?

Vosk has models for English and more than twenty other languages, one model per server. Google Speech supports many more languages and dialects.

Stuck, or would rather not do this by hand?

Deploy it in one click

A private Jitsi server in your own AWS account with SSL, your domain and optional recording, transcription and JWT. Free 15 minute trial.

Start free trial

Talk to a Jitsi engineer

Setup, fixes, branding, recording, scaling. Tell us what is happening and we reply with a plan and a quote.

Get expert help

Related

Free self-hosted Jitsi transcription with Jigasi and Vosk

Add live speech-to-text to self-hosted Jitsi Meet with Jigasi and a local Vosk server. The three settings the stock Docker files get wrong, and how to fix them.

Jitsi transcription or subtitles not working

Fix Jitsi transcription with Jigasi: service-unavailable (SIP mode), SASL not-authorized (no transcriber password) and empty transcripts (Vosk URL).

Jitsi recording setup service (Jibri)

We add reliable meeting recording to your Jitsi server with Jibri: kernel setup, dedicated recorders, concurrent recordings and automatic upload to S3 or your storage.

Jitsi Meet server requirements

How much CPU, RAM and bandwidth a Jitsi Meet server needs, which ports to open, and what recording or transcription adds. Sizing tables from real deployments.