Speech to text API pricing
Billed per second of audio, no minimum. About the API
| Features | What you get | Per hour |
|---|---|---|
| Text | Text with timestamps | $0.12 |
| + Speakers | + who spoke: A, B, C | $0.15 |
| + Names | + their names, where said aloud | $0.20 |
What a month costs
Pick your hours and features. The same hours are priced at four other vendors below.
$20.00 a month
Covered by one $20 pack. That is $0.20 an hour.
| Vendor | Plan | Per hour | Per month | Source |
|---|---|---|---|---|
| MeetWave | API | $0.20 | $20.00 | — |
| AssemblyAI | Universal-2 | $0.19 | $19.00 | Pricing page · checked Oct 21, 2026 |
| Deepgram | Nova-3 | $0.26no names | $26.00 | Pricing page · checked Oct 21, 2026 |
| Speechmatics | Melia 1 | $0.24no names | $24.00 | Pricing page · checked Oct 21, 2026 |
| Soniox | Async | $0.10no names | $10.00 | Pricing page · checked Oct 21, 2026 |
Pay-as-you-go list prices, volume discounts not included. A grey price marked “no names” is that vendor’s price with speakers: it lists no speaker names.
API usage credits
Top up with a pack; larger packs come with bonus credits. Paid credits never expire.
| You pay | Bonus credits | Credits granted | Per hour with names |
|---|---|---|---|
| $5.00 | — | $5.00 | $0.200 |
| $10.00 | — | $10.00 | $0.200 |
| $20.00 | — | $20.00 | $0.200 |
| $50.00 | +10 % | $55.00 | $0.182 |
| $100.00 | +20 % | $120.00 | $0.167 |
| $500.00 | +25 % | $625.00 | $0.160 |
Prices are before tax. Your first key comes with one hour of audio with names, free; it expires 60 days after it is granted.
Pricing FAQ
Per second of audio, rounded up to the next whole second, at the hourly price of the features you ask for. There is no minimum per file and no monthly fee.
No. A job that fails releases what it reserved and is not billed.
Credits you pay for never expire. The free hour on your first key expires 60 days after it is granted.
Bonus credits are not refundable. If a top-up is refunded, you get back at most what you paid, and the bonus that came with it is taken off the balance.
No. Prices are before tax; tax is added at checkout where it applies.
Speaker labels (diarization) tell the voices apart: A, B, C. Names go one step further and put the real name on a voice when the conversation says it, for example when someone is greeted or introduces themselves. When no name can be found, the label stays.
