One rate. Everything included.

One price for everyone, in the app and through the API. Diarization, speaker naming, summaries, translation into any language, voices, broadcasting and sharing are included in every plan.

Pay as you go

Top up credit
as needed

Start free

$5 one-time signup credit, no card required
1 concurrent session
Every feature included
Self-serve help from the docs

Premium

$40

per month

Billed annually

$50 of credit every month+$10 bonus
5 concurrent sessions
Mostly warm server start
Email support

Enterprise

custom pricing

Talk to us

Custom credit and volume rates
Unlimited concurrent sessions
Dedicated engine pool, always warm
Account manager and custom SLA

Pay as you go

Start free and top up credit as needed

Start free

Premium

Ideal for power users and developers

Enterprise

Volume rates and dedicated infrastructure

Talk to us

	$0Top up credit as needed No card required	$40 per month Billed annually	custom pricing

PricingThe same rates in every plan, in the app and through the API. Billed per second.

Included credits	$5 one-time signup credit	$50 of credit every month(333 file hours or 100 live hours)	Custom
Files	$0.15 per hour	$0.15 per hour	Volume rates
Live	$0.50 per hour	$0.50 per hour	Volume rates
When credit runs out	Stops immediately	Grace period	Grace period
No extra cost for translations, summaries, voices, broadcasting and sharing

CapacityMore capacity and faster starts on higher plans.

Start-up speed	Mostly cold	Mostly warm	Always warm
Concurrent transcriptions	1	5	Unlimited
Max file upload size	2 GB	5 GB	Custom
Max live recording length	5 hours	10 hours	24 hours
Dedicated engine pool

RetentionHow long each part of a transcription is kept on file. You can delete anything sooner.

Source audio	7 days	Unlimited	Unlimited
Voices	7 days	Unlimited	Unlimited
Transcriptions	30 days	Unlimited	Unlimited

Support

Documentation
Email support
Account manager
Custom SLA

DevelopersThe API does everything the Dashboard does, from your own code. Included in every plan, billed at the same rate.

API access
MCP server
SDKs (TypeScript, Python)
API rate limit	10 transcriptions per hour	100 transcriptions per hour	Custom

VocalStack compared

Per hour of audio, list prices from each vendor’s pricing page, checked October 2026. If you believe any of this is inaccurate, please contact us.

Translated into

1 language

+$Paid add-onCaveat, hover for details

Model and base price				Features, per audio hour
Provider / model				Audience		Voice		Speakers		Post Processing
Provider / model				Translation	Broadcasting	TTS	Cloning	Diarization	Labeling	Summary	Paragraphs	Titles
VocalStack Live Transcription	High	100	$0.50	Included	Included	Included	Included	Included	Included	Included	Included	Included	$0.50
AssemblyAI Universal-Streaming	Low	6	$0.15					+$0.12					$0.27
AssemblyAI Universal-3.6 Pro Realtime	High	32	$0.45					+$0.12		~+$0.02			$0.59
Deepgram Nova-3 monolingual streaming	Medium	63	$0.46			+$0.27		+$0.12					$0.58
Deepgram Nova-3 multilingual streaming	Medium	10	$0.55			+$0.27		+$0.12					$0.67
OpenAI gpt-live-transcribe	Medium	—	$1.02	+$2.04		+$0.135							$1.02
Gladia Solaria-1 Realtime (Starter)	Low	100	$0.75	Included				Included		Included			$0.75
Speechmatics Linden 1	High	55	$0.30	+$0.65		+$0.099		Included					$0.95
Speechmatics Standard realtime	Low	55	$0.45	+$0.65		+$0.099		Included					$1.10
Speechmatics Enhanced realtime	Medium	55	$0.80	+$0.65		+$0.099		Included					$1.45
Google Chirp 3 streaming	Medium	111	$0.96			+$0.27							$0.96
AWS Transcribe streaming	Medium	99	$0.60			$0.036–0.90		Included					$0.60

Broadcasting

Frequently Asked Questions

: Fifty cents. A live transcription costs $0.50 an hour on every plan, and that one price covers everything that happens on top of it: translation into up to 10 languages, spoken translations, and every listener who opens the link.
Listeners are never counted. Whether 5 people or 500 read along on their own phones, each in their own language, the bill is for one hour of audio. The same is true for languages: adding Japanese to a session already running in Spanish and French does not change the price.
Billing is per second, so a shorter session costs less. Drag the slider to see what live audio of a given length costs:
30 minutes
File, any plan
$0.0750
Live transcription, any plan
$0.2500
: No. A cloned voice reads the translated transcription out loud in that speaker’s own voice, in any language. Thirty seconds of clean audio is enough to make one. Once a voice exists, the speaker is also recognised and labelled by name in future recordings automatically.
When a recording ends, the summary is written for you, and the transcript can be exported as PDF with timestamps and speaker labels, or as Excel with one row per segment. Corrections made after the event are saved to the same transcript. None of this is metered separately; the only thing billed is the audio itself.
: A live transcription has to keep up with the speaker for the whole event, so words reach listeners moments after they are spoken, in every language at once. That work runs continuously, even through pauses and applause.
Files have no audience waiting. They are processed several times faster than real time, in parallel, which is why an hour of recorded audio costs $0.15 instead of $0.50. If nobody needs to read along as it happens, upload the recording afterwards and pay the file rate.
: No. Every live transcription can run in Auto, Speed or Accuracy mode, and all three are billed at the same $0.50 an hour.
Speed puts words in front of listeners as quickly as possible. Accuracy puts the most precise transcription first. Auto balances the two for you as the event unfolds. Pick whichever suits the room; the mode is a quality choice, not a pricing tier.
: Yes. A broadcast is one public link, and any number of live transcriptions can be attached to it: a microphone on stage, a radio stream, an HLS or DASH feed, or a YouTube live. Listeners open the same link and pick their language once.
Each attached transcription is billed for its own minutes of audio. The broadcast itself, the link, the QR code and everyone reading it cost nothing. A private broadcast, protected by a password, is billed the same way.
: On Pay as you go, a live transcription stops as soon as the balance reaches zero, and listeners see the transcript end where the credit did. Add credit to start again; the broadcast link stays the same, so nobody has to rejoin.
On Premium and Enterprise there is a grace period. The transcription keeps running so the event is not cut off, and the extra minutes are billed afterwards at the same rate. Nothing about the session changes for the speaker or the audience.
: On Pay as you go, transcriptions are kept for 30 days, and source audio and cloned voices for 7 days. Export anything you want to keep before then.
On Premium and Enterprise everything is kept until you delete it. Deleting a recording removes both the audio and its transcript, for good.
: No. The API, MCP server and SDKs are included in every plan and billed at exactly the same rates as the Dashboard: $0.15 an hour for files and $0.50 an hour for live. Plan credit applies to API usage too.
Talk to sales about Enterprise when you need volume rates, unlimited concurrent transcriptions, 24-hour live recordings, a dedicated engine pool, always-warm starts, an account manager or a custom SLA.