Extract Insights From PDFs Through AI Conversations

PDF analysis often consumes hours of manual work. Extracting specific data from dense reports, academic papers, or legal contracts requires sifting through pages, cross-referencing details, and organizing findings. You might need tools that accelerate this process without compromising accuracy.

Auntie PDF addresses these by combining Mistral OCR with AI-driven analysis. This free and open-source web application parses PDFs of any size, identifies key information, and enables natural language interactions with documents.

Features

PDF Parsing: You can upload and analyze PDFs of any size. AI can handle everything from short reports to extensive documents.
Intelligent Insights: Receive clear, actionable information and concise summaries.
Chat Interface: You have the ability to ask questions directly about your document. Receive immediate, specific answers.
Powered by Mistral OCR: The world’s best document understanding API from Mistral OCR is used to extract text with accuracy. Auntie PDF understands complex layouts, tables, and various formats.
Self-Hosting Option: You can deploy your own instance of Auntie PDF for greater control over data and customization.

Use Cases

Research: Quickly extract key findings, data, and citations from extensive academic papers. You save hours of manual review.
Legal Document Review: Locate specific clauses, terms, and conditions within lengthy contracts.
Customer Service: Create searchable knowledge bases. Extract key details from customer feedback forms or reports.
Historical Preservation: Digitize archives and make them accessible.
Student Use: Extract important passage in a textbook PDF and chat with it.

Live Demo

How To Use It

1. Visit the Auntie PDF website at auntiepdf.com.

2. Upload your PDF document or enter the PDF URL.

3. Click the “Let Auntie Read It!” button and wait briefly while the AI processes your document. The system will extract and organize all text content from your PDF.

4. Once processing is complete, use the chat interface to ask specific questions about your document. Type your question in the text field and press Enter or click the send button.

Self-Hosting Auntie PDF

You can run Auntie PDF locally on your own machine. This gives you more control over your data and removes the file size limitations of the hosted version.

1. Prerequisites. Before you begin, make sure you have the following installed:

Node.js: Auntie PDF is built with Node.js. You’ll need a recent version (v14 or later is recommended). You can download it from nodejs.org.
npm, yarn, pnpm, or bun: These are package managers for Node.js. Choose your preferred one. npm usually comes bundled with Node.js.
Git: You’ll use Git to clone the Auntie PDF repository. You can get it from git-scm.com.
A Mistral API Key: This is crucial for the OCR functionality. Obtain one by signing up at the Mistral AI platform.

2. Clone the Repository. Open your terminal or command prompt and run the following command to clone the Auntie PDF repository from GitHub:

git clone https://github.com/btahir/auntie-pdf.git

3. Navigate to the Project Directory. Once the repository is cloned, navigate into the newly created project directory:

cd auntiepdf

4. Install Dependencies

Inside the project directory, install the necessary dependencies using your chosen package manager:

npm install
# or
yarn install
# or
pnpm install
# or
bun install

This command reads the package.json file and downloads all the required libraries and packages for Auntie PDF to function.

5. Create the Environment Variable File. Create a new file named .env.local in the root directory of the project (same level as package.json).

6. Add Your Mistral API Key. Open the .env.local file in a text editor and add the following line, replacing "your_mistral_api_key" with your actual Mistral API key:

MISTRAL_API_KEY="your_mistral_api_key"

Important: Do not include any spaces around the = sign.

7. Run the Development Server. Start the local development server with one of the following commands:

npm run dev
# or
yarn dev
# or
pnpm dev
# or
bun dev

This command starts a local server, usually on port 3000. You’ll see output in your terminal indicating that the server is running.

8. Access Auntie PDF Locally. Open your web browser and go to http://localhost:3000. You should now see the Auntie PDF application running locally. You can now upload PDFs without the 4.5MB file size limit imposed by the Vercel deployment.

9. (Optional) Building for Production. If you are self-hosting on your own server, and want to build a production-ready version of the app, do:

npm run build
# or
yarn build
# or
pnpm build
# or
bun build

This creates an optimized out folder.

Troubleshooting

If the server doesn’t start: Double-check that you have Node.js and your chosen package manager installed correctly. Ensure that all dependencies are installed by running npm install (or your preferred package manager’s equivalent) again.
If the OCR doesn’t work: Verify that your Mistral API key is correctly entered in the .env.local file. Make sure there are no typos or extra spaces.
Port Conflicts: If port 3000 is already in use, Next.js will usually try the next available port (3001, 3002, etc.). Check the terminal output for the actual port being used. You can also specify a different port using the -p flag, e.g., npm run dev -p 4000.

Pros

Zero cost for basic document processing
No login or data retention policies
Real-time citation tracking for verifiable answers

Cons

4.5MB file limit on public deployment
Limited batch processing capabilities

FAQs

Q: How does Auntie PDF handle sensitive or confidential documents?
A: When using the public web version, documents are processed through secure channels but are not permanently stored. For highly sensitive information, the self-hosted option provides additional security control.

Q: Can Auntie PDF extract information from scanned PDFs?
A: Yes, Auntie PDF uses Mistral OCR technology which can process scanned documents, though the accuracy depends on the scan quality. Clear, high-resolution scans yield the best results.

Q: How accurate is the information extraction from complex PDFs with tables and graphs?
A: Mistral OCR excels at extracting text from complex layouts and can recognize table structures. However, visual elements like graphs are interpreted based on any accompanying text rather than the visual data itself.

Q: Does Auntie PDF work with PDFs in languages other than English?
A: Yes, Mistral OCR supports multiple languages and scripts, making Auntie PDF suitable for processing multilingual documents.

Q: How can I process files larger than the 4.5MB limit on the web version?
A: For larger files, you can either use the self-hosted version of Auntie PDF which bypasses these limitations or consider splitting your PDF into smaller sections before uploading.