# PageIndex MCP

This MCP server lets an agent search long PDFs through document structure and reasoning. It works with local and online files.

- MCP server · `type=mcp` · Claude Code, Codex, Cursor, GitHub Copilot, Windsurf (Devin Desktop), Cline, Zed · ★ 404
- Author: VectifyAI
- License: MIT · policy=link-only · description only; the file stays with the author
- Source: https://github.com/VectifyAI/pageindex-mcp
- Updated: 2026-07-25
- HTML: https://affpapa.org/en/nevibecode/mcp/pageindex-mcp

## What it does

PageIndex MCP gives an agent a way to work with long PDFs. It builds a hierarchical document tree, then uses reasoning and tree search to locate relevant passages. The approach keeps the document structure and context, without a vector database, chunking, or a fixed Top-K setting.

It supports local and online PDFs, document conversations, and up to 1000 free pages. An API setup needs a PageIndex API key and an MCP configuration using `https://api.pageindex.ai/mcp`; PageIndex App users can connect through OAuth. Local PDF uploads require Node.js ≥18.0.0, so the local server will not start without that runtime.

In Claude Code, add the server with `claude mcp add pageindex-mcp -- npx -y @pageindex/mcp`. This uses `npx` to run the local MCP server and lets the agent work with a PDF available on the machine.

## Requirements
- Needs: Node.js (commands run via npx)

## Install

### Claude Code (native platform)

**Platform CLI**
```bash
claude mcp add pageindex-mcp -- npx -y @pageindex/mcp
```

### Codex

**Platform CLI**
```bash
codex mcp add pageindex-mcp -- npx -y @pageindex/mcp
```

### Cursor

**Config file**
File `.cursor/mcp.json`:
```json
{
    "mcpServers": {
        "pageindex-mcp": {
            "command": "npx",
            "args": [
                "-y",
                "@pageindex/mcp"
            ]
        }
    }
}
```

### GitHub Copilot

**Config file**
File `.vscode/mcp.json`:
```json
{
    "servers": {
        "pageindex-mcp": {
            "type": "stdio",
            "command": "npx",
            "args": [
                "-y",
                "@pageindex/mcp"
            ]
        }
    }
}
```

### Gemini CLI

**Platform CLI**
```bash
gemini mcp add pageindex-mcp npx -y @pageindex/mcp
```

## When to use it
- search a long PDF for a specific requirement
- ask questions about a local PDF
- retrieve related passages from an online file
- trace how a result was found in the document tree

## Example prompts
- Analyze this PDF and find the section about API requirements.
- Compare the conditions in the `Pricing` and `Limits` sections of this document.
- Find every place where authorization is described in this online PDF and explain how the passages relate.

## Similar
- [GitHub](https://affpapa.org/en/nevibecode/mcp/github.md) — MCP server · ★ 33918
- [Netdata](https://affpapa.org/en/nevibecode/mcp/netdata.md) — MCP server · ★ 81115
- [browser-use](https://affpapa.org/en/nevibecode/mcp/browser-use.md) — MCP server · ★ 119160
- [Codebase Memory](https://affpapa.org/en/nevibecode/mcp/codebase-memory.md) — MCP server · ★ 47392
- [mcp-server](https://affpapa.org/en/nevibecode/mcp/mcp-server.md) — MCP server · ★ 43872
- [AIHawk](https://affpapa.org/en/nevibecode/mcp/aihawk.md) — MCP server · ★ 32221

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md