Deployment Models

Last updated 10/20/2021

Access speech recognition models via APIs without deep data science expertise. Run state-of-the-art Deep Learning models in the cloud or on-premiseswherever your business needs.

Deepgram offers the following deployment models:

Deepgram HostedA multi-tenant service running on Deepgram's cloud.
On PremiseA dedicated service deployed to customer-requisitioned cloud instances or customer data centers.

Operational differences

OperationDeepgram Hosted (Multi-tenant)On Premise
Deployment LocationDeepgram’s infrastructureCustomer-requisitioned cloud instance
Customer data center
Infrastructure & Backup ResponsibilityDeepgramCustomer
UpdatesAutomatic rolling updatesCustomer-initiated rolling updates. Excludes critical updates (e.g., security patches), which will be applied as soon as possible.
Service & Uptime ReportingMonitored by DeepgramMonitored by customer