Text-to-speech Limits for Cloud Providers

Limits change from time to time however it is good to know what the limits are (were)

Microsoft

  • Max Audio length produced per turn 10 min/64kB

  • Functional testing example: 4 speech nodes in a row containing 7200 characters each - You need to wait a few seconds for processing but it works well overall.

Amazon (Polly)

  • Allows synthesis of 3000 characters as maximum for a speech node

Google

  • Allows synthesis of 5000 characters as maximum for a speech node

Last updated