What's New in Syncfusion^® Document SDK

AI solutions

AI Agent Tools

Syncfusion Document SDK AI Agent Tools is a .NET library that gives AI agents real, deterministic document processing capabilities. Instead of relying on large language models to guess document structures or generate fragile glue code, your agents can now call production-ready tools powered by the Syncfusion Document SDK. No document processing logic is required in your host application.

This library is available via the Syncfusion.DocumentSDK.AI.AgentTools NuGet package. It’s compatible with .NET 8.0, 9.0, and 10.0.

Key features:

• Comprehensive built-in tools: Prebuilt, AI-callable tools covering the full document processing spectrum-PDF (create, redact, sign, OCR, encrypt, merge, compress), Word (mail merge, track changes, compare, convert), Excel (charts, pivot tables, formulas, export), PowerPoint (extract, merge, split, convert), Office-to-PDF conversion, and smart data extraction (tables, forms, and key-value pairs from PDFs and scanned images as structured JSON).

• Flexible execution architecture: Choose in-memory mode for low-latency execution with automatic cleanup or storage mode with Azure Blob Storage and Amazon S3 support for stateless, horizontally scalable workflows.

• Fully extensible: Extend the library with your own custom AI agent tools by cloning the open-source repository. Custom tools register alongside built-in tools and are callable by the AI agent in the same way, with no framework changes required.

• Enterprise-ready: Compatible with OpenAI, Azure OpenAI, Claude, and custom orchestration systems. The tools run fully on-premises or in secure cloud environments, so no documents ever leave your infrastructure. For practical scenarios, refer to the real-world use cases.

Agent Skills

Syncfusion Document SDK Agent Skills is a knowledge package that grounds your AI assistant in accurate Syncfusion Document SDK APIs, document processing patterns, and platform-specific configurations. Eliminate AI hallucinations and get production-ready document processing code instantly, directly within your IDE.

Agent Skills works with GitHub Copilot, Visual Studio Code, Code Studio, Cursor, Claude, and more.

Key features:

• Broad library coverage: Supports Word (DocIO), PDF, Excel (XlsIO), PowerPoint, Markdown, smart data extraction, and calculate libraries for .NET, Java, Flutter, and JavaScript platforms.

• Two operating modes: Supports Generate code and csx script modes for flexible document generation workflows.
  1. Generate code (mode 1): Triggered by prompts like “how to,” “snippet,” or “example,” it detects your app type, recommends the appropriate NuGet packages, and produces C# code that can be copied and pasted.
  2. Execute via csx script (mode 2): Triggered by action-based prompts like “create,” “generate,” or “open a file,” it writes and runs a temporary .csx script to generate the actual document on your machine.

• One-command install: Get started quickly with npx skills add syncfusion/document-sdk-skills.

• Granular skill selection: Install all skills at once or choose only the libraries you need (e.g., –skill syncfusion-dotnet-pdf).

• Auto-loaded by supported agents: Once installed, skills are automatically detected and loaded automatically with no additional configuration required.

New Control

.NET Markdown Library

This powerful and lightweight .NET library enables developers to create, read, and edit Markdown documents programmatically. Built with a structured document object model (DOM), it allows precise control over document composition while supporting a wide range of Markdown elements and formatting features. Designed for modern .NET applications, it works without external dependencies and helps streamline document generation workflows.

Key features:

• Document creation with DOM: Create Markdown documents programmatically using a structured and intuitive document object model for better control and maintainability.

• Rich content support: Add and organize headings, paragraphs, block quotes, thematic breaks, tables, lists, and checkboxes.

• Advanced text formatting: Apply rich formatting including bold, italic, strikethrough, inline code, subscript, and superscript.

• Media and linking: Insert images and hyperlinks to enhance document content and interactivity.

• Code block support: Easily include formatted code blocks for technical documentation and developer-focused content.

Supported conversion:

Seamlessly convert Markdown to and from multiple popular file formats using Syncfusion libraries:

• Word documents: Convert Markdown to Word and text formats such as .doc, .docx, .xml, .html, and .rtf, and covert Word documents to Markdown.
• Excel documents: Convert Excel files to and from Markdown.
• PowerPoint presentations: Convert PPTX files to Markdown and convert Markdown into PPTX presentations.
• PDF documents: Convert Markdown to PDF format and convert PDF documents to Markdown.

Common

Production-ready Library

The below libraries has been developed to meet industry standards and are now marked as production-ready for Data Extraction:

• Smart Data Extractor: A deterministic C# library that extracts complete document structures from PDFs and images. Designed for .NET workflows, it analyzes visual layout lines, boxes, labels, and alignment to locate and extract elements such as table structure, text elements, images, headers, footers, and form fields with per-field confidence scores for immediate review, export, or integration.
• Smart Table Extractor: A focused, high-accuracy C# library for detecting and extracting tabular data from scanned images and PDFs. It detects table regions, header rows, columns, and cell spans (merged cells) and provides per-cell confidence scores and structured exports ready for downstream processing.
• Smart Form Recognizer: A deterministic C# library for .NET that reliably detects form data in PDFs and scanned images. By analyzing visual layout cues—lines, boxes and circles—it locates form regions and extracts common controls (text fields, checkboxes, radio buttons, signatures), producing clean JSON with per-field confidence scores. Where form fields are detected, the library can also produce a fillable PDF with corresponding form fields added for immediate use.

JavaScript PDF Library

Text element support

PDF text elements enable flexible, layout-aware text rendering in PDFs with support for fonts, styling, alignment, and automatic pagination. They allow precise text placement using coordinates or bounds while flowing content across pages.

Ideal for generating dynamic documents such as invoices, reports, forms, and newspaper-style multi-column layouts where content length varies. It ensures consistent formatting, precise alignment, and smooth text flow across pages or columns without manual pagination or layout adjustments.

See how this feature behaves:

Header and footer templates

Use templates to configure reusable content in the headers and footers of all PDF pages, with reserved layout space to prevent overlap. Include static elements like text, images, and shapes, along with dynamic fields such as page numbers, dates, and times.

This functionality is essential when creating multi-page documents such as reports, contracts, or books that require consistent branding, document titles, page numbering across every page without manually repeating content.

See how this functionality behaves:

XMP metadata support

Embed structured, RDF-based XML metadata (title, author, keywords, dates, rights) within PDFs to make them easier to find, improve interoperability, and boost consistency across tools.

Ideal for document management and archival workflows where PDFs need to include standardized metadata (such as title, author, and keywords). This metadata can then be used by external systems to enable efficient indexing, searching, categorization, and cross-platform interoperability.

.NET PDF Library

HTML-to-PDF conversion enhancement

The Blink binaries used in HTML-to-PDF conversion have been updated to version 147.0.7727.56, improving security by addressing known vulnerabilities and ensuring a safer conversion process.

.NET Excel Library

Place images in cells

The Syncfusion Excel Library lets users place images directly inside cells, where they automatically move, resize, and align with cell changes. This feature works well for structured data scenarios like product catalogs or reports, ensuring images remain properly aligned during sorting, filtering, or resizing.

Excel to Markdown

The Syncfusion Excel Library supports Excel-to-Markdown conversion, enabling seamless content transformation for documentation and web publishing. This feature extracts cell values, images, and cell formatting into clean, readable Markdown while preserving the Excel layout. It can be used to publish Excel-based reports directly to documentation platforms or websites without manual formatting.

Markdown to Excel

The Syncfusion Excel Library supports Markdown-to-Excel conversion, enabling seamless content transformation for data management. This feature easily maps headings, lists, and content into organized Excel data for data analysis, reporting, or collaborative editing.

.NET Word Library

Performance improvement

The Syncfusion Word Library now delivers significantly faster clone and merge operations, achieving a performance improvement of up to 70%.

HTML parsing enhancement

The Syncfusion Word Library now provides improved HTML parsing with comprehensive support for all attributes inside <p> and <h1>–<h7> tags across all selectors, in alignment with Microsoft Word’s capabilities.

.NET PowerPoint Library

PPTX to Markdown

Introducing support for converting PowerPoint presentations(PPTX) into Markdown format, enabling seamless content transformation for documentation and web publishing. Extract slides, text, and structured elements into clean, readable Markdown while preserving the logical flow of the presentation.

Markdown to PPTX

Introducing support for converting Markdown documents into PowerPoint presentations(PPTX), enabling seamless transformation of structured content into visually engaging slides. Easily map headings, lists, and content into slide layouts while preserving document structure.

.NET Data Extraction Library

Customize image saving

This enhancement allows users to access image data as a stream, giving full control over how images are saved in the Markdown file (either as base64 or as a URL). This is especially useful when the application saves Markdown with images by uploading them to the cloud server storage and referencing the same URL in the Markdown.

Configurable OCR settings

This feature allows users to configure OCR settings before extraction. Users can now specify options such as language, Tesseract version, image enhancement mode, and other OCR-related parameters, providing greater control and improved accuracy for text extraction workflows.