Skip to content

docs(devops-guide): add bridge-based transcription guide#645

Merged
damencho merged 5 commits into
masterfrom
transcription
Jul 9, 2026
Merged

docs(devops-guide): add bridge-based transcription guide#645
damencho merged 5 commits into
masterfrom
transcription

Conversation

@bgrozev

@bgrozev bgrozev commented Jun 16, 2026

Copy link
Copy Markdown
Member

Adds a devops-guide page documenting how to enable transcriptions using the bridge-based transcription proxy (opus-transcriber-proxy).

Covers:

  • Running the service — standalone Docker container and Cloudflare (Container + Worker), with an OpenAI example and a note that custom OpenAI-compatible (e.g. self-hosted) endpoints are supported.
  • Prosody — a sample module that forces asyncTranscription=true for all rooms, plus how to enable it.
  • Jicofojicofo.transcription.* config (url-template, http-headers, ping), including Cloudflare Zero Trust auth via CF-Access-Client-Id / CF-Access-Client-Secret headers.

Also notes that Jigasi-based transcription is deprecated and will be removed in a future release.

Detailed service configuration is referenced from the project's own docs rather than duplicated.

Document enabling transcriptions via the bridge-based transcription
proxy (opus-transcriber-proxy): running it as a standalone container
and on Cloudflare, a sample Prosody module forcing asyncTranscription
for all rooms, and Jicofo configuration including Cloudflare Zero Trust
auth via HTTP headers. Note that Jigasi-based transcription is
deprecated.
bgrozev added 2 commits July 7, 2026 09:35
opus-transcriber-proxy now pulls libopus in as a git submodule and no
longer uses the Emscripten 'npm run configure' setup. Update the
standalone-container and Cloudflare build steps to check out the
submodule and use docker:build / the documented Worker flow, and note
the OPUS_BACKEND runtime selection.
Lead the standalone-container instructions with the prebuilt
jitsi/opus-transcriber-proxy image from Docker Hub instead of building
locally. Build-from-source steps live in the opus-transcriber-proxy
repo and are referenced rather than duplicated.

@damencho damencho left a comment

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It misses that config.js needs changing:

transcription: {
     enabled: true,

Comment thread docs/devops-guide/transcription.md Outdated
Comment thread docs/devops-guide/transcription.md Outdated
Comment thread docs/devops-guide/transcription.md
Comment thread docs/devops-guide/transcription.md Outdated
}
```

Reload prosody (`prosodyctl reload`) afterwards.

@damencho damencho Jul 8, 2026

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It needs restarting prosody as it does not load the module, otherwise.

Comment thread docs/devops-guide/transcription.md Outdated
- Change host port to 9090 to avoid conflict with JVB's default 8080
- Drop non-existent room_metadata module from Prosody config
- Replace prosodyctl reload with systemctl restart prosody
- Add &sendBack=true to Jicofo url-template
- Add systemctl restart jicofo note after Jicofo config
@damencho damencho merged commit ce6b471 into master Jul 9, 2026
1 check passed
@damencho damencho deleted the transcription branch July 9, 2026 10:43
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants