# AI use cases for PageIndex (VectifyAI/PageIndex)

Repo: [VectifyAI/PageIndex](https://github.com/VectifyAI/PageIndex) | Use cases: 3 | Page: https://dedupedai.com/use-cases/vectifyai-pageindex/

PageIndex is a vectorless, reasoning-based RAG system that builds a tree index of long documents and lets an LLM search it like a human reader, with local and cloud SDK modes.

This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL.

- Ask questions of a long PDF without a vector database: https://dedupedai.com/use-cases/vectifyai-pageindex/ask-questions-of-long-pdf-without-vectors/
- Store document indexes in the cloud: https://dedupedai.com/use-cases/vectifyai-pageindex/index-documents-in-pageindex-cloud/
- Compare reasoning-based retrieval with your vector RAG: https://dedupedai.com/use-cases/vectifyai-pageindex/compare-reasoning-rag-with-vector-rag/

## PageIndex tutorial: RAG over a long PDF without a vector database

Repo: [VectifyAI/PageIndex](https://github.com/VectifyAI/PageIndex) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README

Submit a document, build a tree index, and chat with it using reasoning-based retrieval.

**Who it is for:** A financial analyst querying a 200-page annual report.

**The problem:** Chunked vector search breaks context in long documents.

**The outcome:** Answers retrieved by navigating the document's structure.

### Steps

1. Install with `pip install -U pageindex`
2. Set OPENAI_API_KEY
3. Create a PageIndexClient and submit_document("report.pdf")
4. Call client.chat with your question and the doc_id

### Prompt to paste into your agent

```text
Using VectifyAI/PageIndex in local mode, submit <report.pdf> and ask <question>. Return the answer with the page or section it came from and verify it in the PDF.
```

### Caveat

Index building runs an LLM over the document, so long files cost more to index.

Risk, credentials: Local mode sends document content to your LLM provider using your key.

Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/

URL: https://dedupedai.com/use-cases/vectifyai-pageindex/ask-questions-of-long-pdf-without-vectors/

## How to store document indexes in the cloud with PageIndex

Repo: [VectifyAI/PageIndex](https://github.com/VectifyAI/PageIndex) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README

Use cloud mode so indexes are built and stored by PageIndex while you chat with your own model.

**Who it is for:** A team sharing indexed documents across services.

**The problem:** Rebuilding indexes per machine is wasteful.

**The outcome:** Cloud-stored indexes reusable by doc_id.

### Steps

1. Set PAGEINDEX_API_KEY and OPENAI_API_KEY
2. Create PageIndexClient with index set to cloud
3. Submit the document
4. Chat using the returned doc_id

### Prompt to paste into your agent

```text
Submit <document> to PageIndex in cloud mode, then ask <question> from a second script using the same doc_id. Confirm both return consistent answers.
```

### Caveat

Cloud mode stores your documents with PageIndex.

Risk, credentials: The PageIndex key gives access to your stored indexes.

Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/

URL: https://dedupedai.com/use-cases/vectifyai-pageindex/index-documents-in-pageindex-cloud/

## How to compare reasoning-based retrieval with your vector RAG with PageIndex

Repo: [VectifyAI/PageIndex](https://github.com/VectifyAI/PageIndex) | Difficulty: Advanced | Time: 1 hour | Basis: From the README

Run the same questions through PageIndex and your existing pipeline to compare answers.

**Who it is for:** An ML engineer evaluating whether to replace a vector RAG stack.

**The problem:** It is unclear when vectorless retrieval is better.

**The outcome:** A side-by-side answer comparison on your own documents.

### Steps

1. Pick 10 questions with known answers
2. Run them through your vector RAG
3. Run them through PageIndex
4. Score correctness and citations

### Prompt to paste into your agent

```text
Answer <10 questions> about <document> with PageIndex and with my vector RAG, then report which got more correct with verifiable citations.
```

### Caveat

Results vary by document type and model choice.

Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/

URL: https://dedupedai.com/use-cases/vectifyai-pageindex/compare-reasoning-rag-with-vector-rag/