Unity Gateway services have none by default, but you can add limits on queries (QPM) or tokens (TPM) per minute to keep capacity and spending under control.
Read more: Databricks docs
In the Ultra Transcenders books
Each book explains Rate limits in context, with comparison tables and the common traps.
Terms in this definition
- Unity Gateway
Governs AI spending, security and access from one Databricks control plane, spanning agents, MCP servers and models. Only a portion applies to
system.aimodel services, including budgets and usage tracking. - Architecture Definition Document
A key deliverable bringing together the main architecture artifacts across the four domains for every relevant state: baseline, transition and target. It sets out, in qualitative terms, what the architect intends.
- QPM
A Unity Gateway setting that caps how many requests model and MCP services accept each minute.
- TPM
Tokens per minute, the measure for Azure OpenAI quota and deployment rate limits. Going over it produces an HTTP 429 response.
- Capacity
A reserved block of compute for a tenant whose size is fixed by its SKU; Fabric F SKUs express it in capacity units (CUs). Workspaces assigned to it unlock features like Copilot, and from F64 upwards people with free licences can view Power BI content.
- CONTROL
Granting this on a securable gives all other permissions on it too, making it the most powerful SQL permission. At database scope that includes UNMASK and ALTER ANY MASK. Warehouse access through the Admin, Member or Contributor workspace roles carries it.
Related terms
- API Management
Azure's API gateway, where policies like rate limits, quotas, ip-filter and validate-jwt are defined once and apply to every API. Production VNet injection is offered in the Premium tier.
- FrontDoorAccessLog
A Front Door log category that captures each request, including its URI, client IP and status. It helps when studying traffic patterns, for instance to decide on rate limits.
- Platform limits
Ceilings on ITPM, OTPM and QPH (input and output tokens a minute, and queries an hour) built into pay-per-token Foundation Model APIs, where the tightest one wins. Any rate limits of your own are a different thing.