Docling now supports read/write outputs to Amazon S3, Google Cloud Storage, Microsoft Azure Blob Storage, Google Drive, and Microsoft SharePoint, with sources including IBM FileNet, Box, and Databricks Volumes.

The team plans ongoing expansion of deployment options, an improved user experience, a broader connector ecosystem, and smarter document processing to deliver AI-ready data for search, RAG, and AI agents.

PDF processing is faster, up to 3.8x with docling-parse, and a new native PDF pipeline reduces external dependencies, complemented by a backend-no-dependency optimization for better performance.

Downstream integrations now include OpenSearch (and Amazon OpenSearch Service), DataStax AstraDB for embeddings, and Kafka ecosystems (Confluent Cloud and Amazon MSK) to cut custom integration work.

Docling for IBM watsonx is a fully managed document intelligence service that converts unstructured documents into structure-aware data for enterprise search, RAG, and AI agents.

Converted outputs can be exported as JSON Lines chunks or sent directly to supported downstream targets, with chunk-level delivery optimized for RAG and enterprise search efficiency.

New file formats include Outlook .msg, BoxNote, MHTML, RTF, with minimal AFP support and full support for Apple Pages .pages; improvements preserve document hierarchy, reading order, OCR, Office content, and Markdown tables.

Enhanced MCP (Model Context Protocol) integration lets converters access S3, Google Cloud Storage, and Azure Data Lake Storage directly, and reuse cached conversions when content/settings are unchanged; MCP server ranked in the top 0.5% globally as of September 2026.

Since the June 2026 launch, the service has expanded cloud availability, connectors, formats, and production capabilities to simplify preparing enterprise content for AI workflows.

Users can try Docling for IBM watsonx for free via a signup link.

DocLang and DCLX provide AI-native output formats within IBM watsonx.data, with DocLang encoding content/layout in XML-based format and DCLX packaging representations for portable reuse across AI workflows.

Organizations can procure Docling through IBM Marketplace, AWS Marketplace, Microsoft Marketplace, or the Azure portal, with deployments targeting AWS or Microsoft Azure clouds.