ai_prep_search supports plain text and markdown input (Beta)
The ai_prep_search() SQL function now accepts plain text and markdown STRING content in addition to the structured VARIANT output of ai_parse_document . You can now chunk and prepare text that you already have as a string for AI Search indexing and retrieval-augmented generation (RAG) pipelines, without first running it through document parsing.
Read Databricks's release noteshttps://docs.databricks.com/aws/en/release-notes/product/2026/august#ai_prep_search-supports-plain-text-and-markdown-input-beta
Summaries of vendors' own notes. Product names and logos belong to their owners; logos via logo.dev.
More Databricks Data Intelligence Platform releases
Zhipu AI GLM 5.3 now available as a Databricks-hosted model
UpdateAI agentsGovernanceRegions
Restrict service principal OAuth secrets to specific API scopes
UpdateSQLSecurityDeveloper tools
Upcoming end of life for the Supervisor API
DeprecationAI agentsDeveloper tools
Write data back to Databricks from Google Sheets
UpdateGovernanceData integrationObservability
Also shipped on Aug 25, 2026
BigQuery data governance tags are supported in Terraform
PreviewSQLGovernanceDeveloper tools
General availability: Bring Your Own Cloud (BYOC)
GAAI agentsVector searchPricing