pdfAssistant MCP Server
VerifiedFeaturedLocalAbout pdfAssistant MCP Server
A hosted MCP server that lets compatible assistants perform professional PDF operations, from OCR and redaction to conversion and form handling. It connects to clients over Streamable HTTP with OAuth 2.0 and requires a paid pdfAssistant plan.
Key features
- 01Convert files and PDFs between formats
- 02OCR and extract document content
- 03Merge, split, and modify PDFs
- 04Encrypt, sign, watermark, and redact
- 05Inspect PDF metadata and structure
Use cases
- 01Making scanned contracts searchable with OCR
- 02Redacting and encrypting account statements
- 03Converting PDFs to Office or image formats
- 04Merging and compressing document sets
MCP support
Exposes PDF processing operations to MCP clients through a hosted Streamable HTTP endpoint with OAuth 2.0 authentication.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
We haven’t collected a package or remote endpoint for this server. Follow the instructions in the project README.
Tools
Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.
Details
From the project READMEpdfAssistant MCP Server
Professional PDF processing for Claude, ChatGPT, Codex, and any MCP-compatible assistant. Drop in a PDF, describe the result you want, and get a finished file back: merged, split, compressed, OCR'd, redacted, encrypted, watermarked, converted, and more than 40 other professional operations.
| Endpoint | https://chat.pdfassistant.ai/mcp |
| Transport | Streamable HTTP |
| Auth | OAuth 2.0 (browser popup, no keys to paste) |
| Requires | A paid pdfAssistant plan: Plus, Pro, Business, or Enterprise (pricing) |
| By | pdfAssistant.ai, a product of Datalogics, Inc. |
The server is hosted. There is nothing to install or run from this repository; it exists to document the endpoint, setup, and capabilities for MCP registries and client directories.
Why this exists
General-purpose AI assistants can read a document, tell you about it, and generate a basic file. What they lack is a professional PDF engine: OCR on scanned pages, conformant PDF/A and PDF/X output, permanent redaction, digital signatures, form flattening, prepress color and transparency handling. The pdfAssistant MCP server gives the assistant those tools to call on your behalf, so a request like the ones below produces a finished, standards-conformant file rather than a best-effort approximation.
- "OCR this scanned contract so the text is searchable"
- "Merge these three PDFs, then compress the result"
- "Redact the account numbers in this statement and encrypt it"
- "Convert this PDF to Word"
- "Turn this spreadsheet into a PDF/A"
- "Watermark every page with DRAFT and flatten the form fields"
Operations run on the same engine behind pdfRest, built on Adobe PDF Library technology.
Quick start
1. Have a paid pdfAssistant account. Plus, Pro, Business, and Enterprise all include MCP access. The Starter plan does not. Sign up or upgrade.
2. Add the server to your client.
Claude Desktop and Claude Code:
claude mcp add --scope user --transport http pdfAssistant https://chat.pdfassistant.ai/mcp
Then type /mcp in Claude and sign in when the OAuth window opens.
Codex CLI:
codex mcp add pdfAssistant --url https://chat.pdfassistant.ai/mcp
codex mcp login pdfAssistant
ChatGPT (web): turn on Developer Mode in settings, add a remote MCP app pointing at https://chat.pdfassistant.ai/mcp, and complete the OAuth sign-in.
Any other client: the server speaks standard Model Context Protocol over Streamable HTTP with OAuth 2.0, so any client that supports remote MCP servers with OAuth can connect using the endpoint above.
3. Attach a PDF and ask. Full announcement with per-platform setup details: pdfAssistant now plugs directly into Claude, ChatGPT, and more.
What it can do
| Category | Operations |
|---|---|
| Convert | Office, HTML, image, and email files to PDF; live webpage capture; PDF to Word, Excel, PowerPoint, Markdown, JPEG, PNG, TIFF, GIF, BMP; PDF/A and PDF/X; XFA to AcroForms |
| Optimize | Compress, linearize, rasterize, convert colors, flatten transparencies, layers, and annotations |
| Secure | Encrypt, restrict, digitally sign, watermark (text or image), remove encryption, remove restrictions, two-step redaction (preview first, then apply permanently) |
| Extract | OCR to searchable text, extract text, extract images, export form data, unpack ZIP archives |
| Modify | Merge, split, add attachments, images, and text, fill and flatten forms, bundle outputs to ZIP |
| Inspect | Metadata and structure: tags, permissions, fonts, page count, security settings |
Full list with a guide per operation: pdfassistant.ai/blog/pdf-capabilities.
Usage and credits
Each document operation the assistant runs uses 20 Credits from your monthly plan allotment. Monthly Credits reset on your renewal date; Extra Credits can be added at any time and never expire. Plan details and credit math are on the pricing page.
Security and privacy
- Controls independently audited for SOC 2 Type 2 (Trust Center)
- pdfAssistant does not use your documents to train AI models. Document handling by the AI assistant you connect through (Claude, ChatGPT, or another client) is governed by that provider's own policies
- TLS 1.3 in transit, AES-256 at rest
- Files processed on isolated infrastructure and deleted automatically within 24 hours
- OAuth credentials live in the browser session, not in your MCP client config
More ways to use pdfAssistant
- Workflows and Batch Processing: build a multi-step document process once on pdfassistant.ai and run it against thousands of files.
- AI Studio: the web platform's workspace for one-off document tasks in plain language.
- Building your own integration? pdfRest is the REST API that powers pdfAssistant, for direct use in your code.
Support
© Datalogics, Inc. pdfAssistant and pdfRest are products of Datalogics.
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.