AI Integration
A contract that answers questions, an invoice created from a sentence, a policy check that cites its sources: each of these combines a language model with documents. In TX Text Control .NET Server, the model interprets the request, and ServerTextControl reads, creates, edits and exports the documents. The TX Text Control AI preview adds local models, document tools and private knowledge, and older samples show the same ideas with OpenAI's hosted API.
How AI Fits Into a Document Application
- Text for the model.
ServerTextControlextracts the text of a PDF, DOCX or other document, or exports it as Markdown, which keeps headings, lists and tables with fewer tokens. - A model you choose. The model runs on hardware you control with llama.cpp, or at a hosted provider such as OpenAI.
- Tools for documents. The MCP document server offers the document operations of TX Text Control as tools that a model requests and your application runs.
- Knowledge from approved documents. Retrieved passages ground answers and cite their sources.
- Results you verify. Your application checks tool results and lets people review generated content before it counts.
Find the Right Page
| If you want to | Read |
|---|---|
| Understand the roles of model and document engine, the building blocks and the data protection questions | Overview |
| Get to know the packages and samples of the TX Text Control AI preview and add its endpoints to ASP.NET Core | Text Control AI |
| Run models on your own hardware, provide the llama.cpp runtime and choose hardware and model size | Local / Private LLMs |
| Ask a model questions about a PDF or DOCX document and handle long documents | Chat with Documents |
| Classify documents, extract structured data, review contracts or describe images for PDF/UA | Document Analysis |
| Index approved documents and ground answers in them with citations | RAG |
| Host TX Text Control document tools for AI clients and secure them | MCP Document Server |
| Let a model call functions and document tools, and plan and verify document changes | Agent Workflows |
| Connect GitHub Copilot, Cursor or Claude Code to the Text Control MCP Server and install Agent Skills | AI Coding Assistants |
| Find runnable samples and their repositories | Examples |
Where to Start
- Overview: the concepts and the building blocks.
- Chat with Documents: a first console program that answers questions about a PDF with a local model.
- Text Control AI: the complete web sample with Chat, Document Studio and the MCP document server.