Per minute, from zero, without a call.
The floor is the barrier, not the rate. The nearest comparable vendor asks for $135 a month committed annually — $1,620 before you have processed a second of audio; two others will not tell you the price at all. Ours is on this page, the whole ladder down to $0.0005, and it starts at nothing.
One plan, one meter
most teams$0.0010per minute — and that is the ceilingThe first 10,000 minutes each month are free, and the rate falls as you send more. No tier to negotiate onto, no minimum, no call.
- 10,000 processed minutes free every month — not a trial, it does not expire
- $0.0010/min is the highest rate there is: $0.0008 over 100k, $0.00065 over 1M, $0.0005 over 10M
- A third under the metered alternative — Krisp VIVA on Pipecat Cloud is $0.0015/min after the same 10,000 free
- All four models, no feature gating, no enterprise edition
- Instant key — no card to start, no demo call, no annual commitment
- Split-pipeline routing and shadow scoring included
- Hard concurrency cap instead of surprise overage billing
| Vendor | Published terms | At 25,000 min |
|---|---|---|
| Anecho | 10,000 free, then $0.0010/min falling to $0.0005, no floor | $15 |
| ai-coustics | $135/mo floor, committed annually; $0.00135 at 100k | $135 |
| LiveKit — voice isolation | $0.0012/min, metered since 1 May 2026, platform only | $30 |
| Pipecat Cloud — Krisp VIVA | 10,000 free minutes, then $0.0015/min, platform only | $23 |
| Telnyx — Krisp | $0.005 per direction, platform only | $125 |
| Krisp SDK direct | $0.0015/min — the same VIVA rate Pipecat Cloud resells, taken direct | $38 |
| Picovoice Koala | Unpublished — pricing page removed | unpublished |
Comparison figures are the vendors’ own published list prices, re-checked against their pricing pages on 2026-08-13 and normalised to one processed stream per minute. Every row here separates one talker from the rest. The broadband noise suppression the large platforms bundle for free is a different product and is not listed, because putting its price beside ours would compare two numbers that buy different things. Of the engines that do the job, most are gated on adopting a platform, one on an annual commitment, and two do not publish a price at all.
| Minutes per month | Rate per processed minute |
|---|---|
| First 10,000 | $0.00 |
| 10,000 – 100,000 | $0.00100 |
| 100,000 – 1,000,000 | $0.00080 |
| 1,000,000 – 10,000,000 | $0.00065 |
| 10,000,000+ | $0.00050 |
Graduated, not tiered: each band is priced separately and applied automatically as your volume grows. Nobody has to ask for the discount and nobody has to negotiate it — including the bottom rate, which is the number our competitors make you call for.
Every model, on every plan.
There is no tier where the good model is hidden. The paid plans buy volume and deployment options, not access.
Clearline
primary-speaker isolation at 8 kHz, enrollment-free, streaming
Chamber
Wideband (16/48 kHz) speech enhancement + primary-speaker isolation
Onset
Voice activity detection and turn-taking
Nyquist
Call-audio quality scoring and failure prediction
| Line | Price | Why |
|---|---|---|
| Clearline — native 8 kHz + residual echo suppression | Standard ladder, no premium | It is the cheapest configuration we run and the highest-value to the buyer. Putting a surcharge on the wedge is how you fail to drive it in. |
| Chamber — wideband 16/48 kHz enhancement | Standard ladder | 48 kHz is edge-only: hosted, a 48 kHz model costs more per minute than the list price. |
| Onset — VAD and turn-taking | $0, bundled | Silero VAD is MIT and excellent. It is also the branch our own benchmark says enhancement actually helps, so it has to be bundled for the split pipeline to be the default. |
| Nyquist — call-audio scoring | $0.0004 per scored minute | Free under 100,000 scored minutes a month. |
| Null Test Private — the benchmark run on your own audio | $2,500 one-time | It converts precisely because the report sometimes says turn enhancement off. |
The things people ask before signing up.
- What counts as a processed minute?
- Wall-clock audio pushed through a session, summed across streams and rounded up per session to the nearest second. Silence counts, because the model runs on it. Sessions you open and never feed cost nothing.
- What happens when the free minutes run out?
- On a test key, sessions keep running and return the unprocessed signal, so your agent degrades to the audio it would have had without us rather than failing. On a metered key it just bills. We never auto-bill overage on a free account — a runaway loop hits a hard concurrency cap instead of generating an invoice.
- Why is there no monthly floor?
- Because the nearest comparable vendor's floor is $135 a month committed annually — $1,620 before you have processed a second of audio — and that number is what stops a developer trying the product on a Saturday. We would rather have the Saturday.
- Do I have to talk to someone to get the volume rate?
- No. The whole ladder is published and applied automatically, including the $0.0005 rate above 10 million minutes a month. That is the number our competitors make you call for.
- Do you train on my audio?
- No. Metered traffic is processed and discarded; nothing is retained for training. If you need that in writing it is in the DPA, and the on-device and self-hosted options remove the question entirely.
- Is the benchmark part of the paid product?
- No. Null Test — the dataset, the harness and every published result — is open and free, including for people who use it to argue that a competitor beats us. We sell the private run on your own audio, never the public one.
- Does the rate go up at any point?
- No. $0.0010 a minute is the ceiling, not the entry price — it is what you pay on the first billable minute and it only falls from there, automatically, as monthly volume crosses 100k, 1M and 10M minutes. There is no tier to negotiate onto and nothing to ask for.