A Unity Gateway setting that caps how many requests model and MCP services accept each minute.
Also called queries per minute.
Read more: Databricks docs
In the Ultra Transcenders books
Each book explains QPM 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. - MCP
Model Context Protocol, an open standard for making tools and data available to models. To use one, a Foundry agent points at a remote MCP server, naming it with a server label and giving its URL.
Related terms
- Rate limits
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.
- TPM
Tokens per minute: how Unity Gateway measures a rate limit on model services, next to QPM. Go over a limit and you get HTTP 429; MCP services aren't limited by tokens.