On-Premise Guide

Last updated 10/20/2021

Deepgram provides a variety of deployment models, including an on-premise offering, which is an isolated service deployed to customer-requisitioned cloud instances or customer data centers.

If you're an IT administrator deploying an isolated instance of Deepgram speech analytics products, read on for high-level guidance on setup, suggested configuration, periodic maintenance, and frequently asked questions.

This guide covers multiple Deepgram products. If you aren't certain which products your contract includes, please consult a Deepgram sales representative.


Detailed troubleshooting and on-demand support require an on-going support contract with Deepgram. To learn more, please contact a Deepgram sales representative.


This guide is intended to address the customer deployment lifecycle of Deepgram products:

  • Initial deployment
    • Pre-requisites
    • Recommended hardware
    • Recommended system configuration
    • Obtaining Deepgram software and data files
    • Product configuration
    • Testing
  • Periodic maintenance


Throughout this guide, we suggest commands to execute in a shell. These commands are prepended by a prompt character. The prompt character represents the command prompt and should not be typed. If the prompt is a "$" (dollar) symbol, the command is intended to be run as your regular (unprivileged) user. If the prompt is a "#" (hash or pound) symbol, it is intended to be run as the root user.