$0
$10/Month
$100/Month
The main distinction between user levels (or plans) is hourly limits. Using default settings a smart speaker, for example could be on 24/7 and realize a 1:1 relationship between QSS and real time. Note, both STT and TTS contribute to hourly QSS consumption. If you are going to require professional grade STT and TTS where both accuracy and performance are needed as fast as possible, then the $10 monthly package will probably work best. Enterprise is only needed if you anticipate a lot of transcription grade STT with very large wav input files. Note, the data share flag will also affect QSS calculations.
$0
Get Now$10/Month
Get Now$100/Month
Get NowQuality speech seconds represents the amount of resource consumed by an operation. It is the product of wav file length in seconds (at the default rate) and the quality multiplier and represents the cost of an operation in both bandwidth and CPU consumption. See models and tradeoffs for more detail on how QSS is calculated. The system is designed such that a normal voice assistant using the default STT and TTS models should easily be able to operate 24/7/365. When an account exceeds its hourly limits it is simply cut-off until it becomes eligible to consume again.