Every voice,
understood.

Deepgram Nova-3 transcription with GPT translation. Accuracy is highest with a clear mic signal and a single speaker; latency is typically 2–5 seconds end-to-end.

CaptionCast supports 12 languages: English, Spanish, Portuguese, French, German, Italian, Russian, Chinese (Mandarin), Japanese, Korean, Arabic, and Hindi. Every translation pair routes through English — for example, Spanish → French is handled Spanish → English → French. Pick the direction for each meeting on the control panel.

Credits are per caption language. A meeting captioned in one language, or translated into one other language, uses 1 credit for every minute of captioning. Each additional translation language adds 1 more credit per minute:

• English captions plus Spanish: 1 credit per minute
• English into Spanish and French: 2 credits per minute
• English into Spanish, French and Portuguese: 3 credits per minute

The rate is shown on the meeting form before you save it, and again on the control panel before you start, so the cost is never a surprise.

No. CaptionCast runs entirely in the browser — Chrome or Edge recommended. A laptop's built-in mic works; a wired lavalier or a direct PA line-out gives the cleanest captions.

CaptionCast bills by minutes streamed, not by viewer — so price-per-attendee shrinks as your room grows. For a 100-person room:

• A $19 Growth Pack (500 credits — roughly seven weekly services in one language) works out to ~$0.03 / person across those events.
• A $49 Pro Pack (1,500 credits) works out to ~$0.02 / person across a full season of events.

For nonprofits, churches, and accessibility teams that need to justify the line item, that's typically less than the cost of one printed program.

Only time with the mic actively recording counts. Your current balance and usage are visible at /org/usage. A meeting already in progress is allowed to finish, and a new one cannot start until you buy another pack (or turn on auto top-up, which charges your saved card when credits run low) — your existing transcripts and recordings stay accessible either way, and new minutes are available the moment the purchase completes.

Yes. Share the display URL or project the QR code. Attendees open it in any browser — no app install needed. Each viewer can pick their own language (on multi-language meetings), their own font size, and their own theme.

Yes. Audio is recorded by default, together with the transcript. After the meeting ends, open /meetings/:id/playback to play the audio back with the captions synced. Transcripts can also be exported from the meeting detail page, and the summary you land on when a meeting ends links straight to the transcript.

Owners and admins can turn audio recording off for the whole organization, and any meeting can override that. With recording off no audio is kept: there is no audio playback and our support staff cannot review the audio, but the transcript is still saved and captions work exactly the same. The choice is read when a meeting starts.

Audio recordings are deleted automatically 90 days after they're recorded (sooner if your organization sets a shorter retention limit). Our speech-to-text provider, Deepgram, receives the live audio whether or not you record it; its standard terms let it keep that audio and use it to improve its speech models.

Not yet. Viewers use any mobile browser — there's nothing to install. Operators use a laptop or desktop browser (Chrome or Edge) because CaptionCast needs WebAudio access to the mic.