SPEECH-TO-TEXT/ TRANSCRIPTION OFFLINE

$60/Port/Mo

SPEECH-TO-TEXT/ TRANSCRIPTION REALTIME

$60/Port/Mo

SPEECH ANALYTICS API
(INCL. TRANSCRIPTION)

$75/Port/Mo

How do you measure billable time on Voicegain Cloud?

For offline STT/Transcription or speech analytics, the billable time is the duration of the submitted audio file (including any silence). For streaming audio, usage is measured and billable for the entire session. For call center two channel (stereo) audio, we only bill for duration of the entire call/session. We do not bill for each individual channel separately.

Platform usage is subject to a 6 second minimum and then measured in 1 second increments(for both Offline and real-time STT).  So if a session lasts 4 seconds, we bill for 6 seconds. If session is 7 seconds, we bill for 7 seconds.

What does Edge mean? What are server requirements?

Voicegain Edge means the Voicegain Platform - [the ASR  - offline and realtime models, the APIs, MRCP server (if required), Edge console, logging & monitoring system] is deployed either on bare-metal in a datacenter or on Client's VPC with a cloud provider (AWS, Azure, GCP, etc). Voicegain is a deep learning based ASR. So clients need to provide servers or compute instances that are GPU based. Voicegain offers in-depth recommendations of server configurations based on client's use-case

How do you define Port/Session on Voicegain Edge?

Licensing for Voicegain Edge is based on number of concurrent ASR Sessions or Ports. A single port of real-time ASR can process a single minute of audio in one minute. Number of concurrent real-time cannot exceed licensed number of ports. For offline, clients may submit more requests than licensed number of ports. Voicegain queues such requests. Based on hardware provisioned, Voicegain can process audio more than 100 times faster than audio duration. So 1 hour of audio can be processed in 30 seconds. Please contact us for how we calculate number Ports for offline use.