Service Limits for Generative AI

Learn about the limits for dedicated AI cluster resources in OCI Generative AI.

By default, the number of dedicated AI clusters that you get per tenancy is 0. For the name of the dedicated AI cluster that you're requesting an increase for, go to Offered Pretrained Foundational Models in Generative AI, select a model card and see the section about the Dedicated AI cluster for the model. Note that (on-demand only) models don't have a dedicated AI cluster option.

For imported models, see the Resource Request & Pricing.

To request dedicated AI clusters for a tenancy, see Creating a Limit Increase Request.

Project Limits

In a tenancy, you can have up to 50 projects in OCI Generative AI.

Application Limits

The following table list the limits for applications in OCI Generative AI.

Application limits
Limit Default Maximum Can request a service limit increase Limit name
Applications per tenancy 10 50 Yes hosted-application-count
Artifacts per application 20 50 Yes artifacts-per-application-count
Managed storage options per application 3 3 No
Environment variables per application 20 100 Yes environment-variables-per-application-count
Maximum replicas per application 30 50 Yes max-replicas-per-application-count
PE/RCE DNS proxies per tenancy 3 10 Yes pe-rce-dns-proxy-count
Managed storage systems per storage type per tenancy 3 10 Yes managed-storage-per-type-count