Skip to content
Lucas Lessa
Café com Features
Voltar para as notícias
DatabricksPublic PreviewImpacto: Médio

Session metadata and token details added to AI Gateway usage table

The Unity Gateway system table now captures coding-agent metadata, invocation sources, and granular token and search details.

O que mudou

Use cases

Imagine you need to audit the cost and behavior of AI agents and enterprise models using Unity Gateway. With the new fields in the usage table, you can track coding agent sessions, analyze prompt cache usage, and monitor search counts.

  • Monitor sessions from integrated coding agents.
  • Analyze tokens written to 5-minute and 1-hour prompt caches.
  • Track file and web search counts.

How to use

You can query the updated data directly using SQL in Databricks SQL or a notebook.

SELECT * FROM system.ai_gateway.usage;

What happens in this code:

  1. Accesses the Unity Gateway system table system.ai_gateway.usage.
  2. Returns all model and service usage records, including the new session metadata and token detail columns.

Limitations and requirements

  • Feature is in Public Preview.
  • Requires a Databricks workspace in a Unity Gateway supported region.
  • Requires Unity Catalog enabled.
  • By default, account and metastore admin roles are required to view or query the table.
  • Usage tracking is a billed Unity Gateway feature.
Ver nota oficial do Databricks

Usage tracking adds session metadata and token details (Public Preview)

Resumo e explicação gerados por IA a partir da release note oficial. Em caso de dúvida, a documentação oficial da plataforma prevalece.

Relacionadas