The Snowflake Cortex AI_PARSE_DOCUMENT document is now generally available with advanced layout extraction capabilities. This fully managed SQL function extracts the layout of the page in Markdown format, preserving text, tables, and structural elements from documents with enterprise-grade accuracy and scale.
Note
The AI_PARSE_DOCUMENT function is the new version of SNOWFLAKE.CORTEX.PARSE_DOCUMENT. The old function is still supported, but Snowflake recommends using the new function.
Key capabilities of AI_PARSE_DOCUMENT include:
- Complex layout mastery: Accurately process multi-column research papers, financial reports, and technical documentation while preserving reading order and document hierarchy.
- Precise table extraction: Maintains table structure, headers, and relationships from financial statements, regulatory filings, and data-heavy documents for downstream analysis
- Advanced Layout Preservation Handles mixed content including embedded images, pull quotes, and complex formatting without losing context or meaning
For more information, see AISQL AI_PARSE_DOCUMENT.
Source: Snowflake
Latest Posts
- Mistral AI Devstral 2 now available on Amazon Bedrock

- Amazon SageMaker HyperPod now provides comprehensive observability for Restricted Instance Groups

- Cloudflare Tunnel, Cloudflare Tunnel for SASE – Cloudflare Tunnel and Networks API will no longer return deleted resources by default starting December 1, 2025

- Workers AI – New conversion options for Markdown Conversion






