Speech-to-text API pricing

Yes. New accounts get 50€ in free credits, valid for one year. That's roughly 312 hours of async transcription on solaria-3, 185 hours on solaria-1, or 73 hours of real-time. Once used, top up to continue, or get in touch for a demo.

Gladia runs on a prepaid credit wallet. You can top up manually, or set auto top-up to refill when your balance drops below a threshold you choose. You'll get notified on low balance, top-ups, and failed payments, so there are no end-of-month surprises.

Auto top-up adds credits to your wallet when your balance drops below a threshold you set, charging your existing payment method. You can change the threshold or the amount, or turn it off, from your billing settings at any time.

When you top up 1,000€ or more, we add free credits: +10% at 1,000€, +15% at 2,500€, and +20% at 5,000€ and above. Top-up credits never expire.

No. Core features like custom vocabulary are included in every request. Add-ons such as summarization or translation are billed per hour only when you enable them, at the rates listed on this page.

Yes, automatically. Your hourly async rate drops as your monthly volume grows. For guaranteed rates at high volume, contact us.

Yes, at no extra cost. Self-serve is opted in by default, and you can opt out anytime from your profile. Enterprise is opted out by default.

You can sign up for Self-serve directly. To move to Enterprise, or to change an existing contract, contact our sales team.

Self-serve runs on prepaid credits with no subscription, so there's nothing to cancel. The Enterprise plan follows its contract terms.

Both units are in common use, which makes providers hard to compare. Gladia publishes per-hour rates starting at 0.16€ an hour for async transcription (Solaria-3) and 0.68€ an hour for real-time streaming on the Self-serve plan. In per-minute terms that is 0.0027€ and 0.0113€ a minute.

Yes, at every provider. Real-time holds a live connection open and returns results as the audio arrives. On Gladia, real-time is 0.68€ an hour against 0.16€ (Solaria-3) for async transcription on Self-serve. Where a delay of a few minutes is acceptable, async is the cheaper path.

The Self-serve plan allows 30 concurrent real-time sessions and 25 concurrent pre-recorded jobs. Enterprise removes the concurrency limits. Requests above the limit receive a 429 response and can be retried. The concurrency page in the Gladia docs covers queue behaviour and how to request more capacity.