How Much Does Gladia Transcription Cost?
How pricing works
You pay for the duration of audio processed, whatever the source (meetings, calls, podcasts, videos…). There is no subscription and no unused minutes: usage is deducted from a prepaid credit wallet.
Your hourly rate depends on three things:
Mode and model — asynchronous (pre-recorded) transcription is priced per model (Solaria-3 or Solaria-1). Real-time transcription has its own rate.
Add-ons — audio intelligence features such as summarization, sentiment analysis, translation, named entity recognition or PII redaction are billed per hour, only when you enable them.
Monthly volume — asynchronous rates drop automatically as your total monthly volume grows.
Core features (100+ languages, language detection, diarization, custom vocabulary, word timestamps…) are included at no extra cost.
Current rates and cost estimate
Rates, volume tiers, add-on prices and top-up bonuses are kept up to date on our pricing page. It also has a calculator: pick the mode, model, add-ons and monthly volume to get your estimated monthly cost.
New accounts start with €50 in free credits. For annual contracts, custom models, zero data retention or unlimited concurrency, see the Enterprise plan on the same page and contact our sales team from there.
Real-time transcription billing
Real-time is billed on the total duration of audio streamed through the WebSocket, including silence, background noise and empty audio frames. As long as the connection is open and audio is sent, it counts toward the billed duration.
How multi-channel audio is billed
Different content on each channel → each channel is billed as a separate audio stream.
Identical content duplicated across channels → the audio is billed only once.
Have questions?
Contact support from your dashboard, or reach our sales team if you expect high volumes.