Chat with your PDF
100% in your browser

Your file never leaves your device. No sign-up. No upload. AI-powered answers with page citations.

🔒 No upload 🧠 Browser AI (Llama 3) 📄 50 MB max 🆓 Free, no sign-up
📄
Drop your PDF here or click to browse
document.pdf
Detecting...

How It Works

Three simple steps. Everything happens on your device.

1

Drop Your PDF

Drag and drop or browse for a PDF file up to 50 MB. It stays on your device the entire time.

2

AI Reads It Locally

pdf.js extracts the text. The AI model (Llama 3 via WebGPU) runs directly in your browser.

3

Chat & Get Answers

Ask questions in any language. Get AI answers with exact page number citations.

🟢 Browser AI

Llama 3.1 8B / 3.2 3B / 1B via WebGPU. Best privacy. Needs a GPU with 4-6 GB VRAM. First run downloads ~2-4 GB (cached after).

🔵 Your API Key

Bring your own OpenAI or Anthropic key. Calls go directly from your browser. We never see your key or data.

🟡 Text Search

No AI needed. Searches your PDF by keyword and shows matching sections with page numbers. Works everywhere.

Browser AI Tools on This Site

Every tool below runs in your browser. Files stay on your device.

Chat with PDF

Ask questions about a PDF and get answers with page citations.

PDF Summarizer

One-click summary of a PDF.

PDF Translator

Page-by-page translation (EN, KO, JA, ZH, ES, FR, DE).

Contract Analyzer

Flags risky clauses, unusual terms and missing protections.

Resume Analyzer

Scores, strengths, weaknesses and suggestions for a resume PDF.

Chat with Excel

Chat with Excel or CSV data in your browser.

Excel Auto Analyzer

Statistics and charts for a spreadsheet.

CSV Analyzer

Chat with a CSV file and get insights.

Local GPT

General AI chat in your browser, with optional PDF, DOCX or spreadsheet context.

Chat with Word

Ask questions about a DOCX file.

Code Explainer

Paste code and get an explanation.

Email Writer

Draft an email with AI.

Chat with Excel and Local GPT used to live at chatwithexcel.tools and localgpt.tools. They moved here on 2026-10-13. Settings saved in your browser at the old addresses (model choice, API key, theme) and AI model files already downloaded there do not carry over, so the model downloads once more here. More browser tools: Tools That Don't Upload.

Verify It Yourself

Don't trust us? Check it yourself in 10 seconds.

1

Open DevTools

Press F12 or Ctrl+Shift+I in your browser.

2

Go to Network Tab

Filter by "Fetch/XHR" or search for POST requests.

3

Drop a PDF & Chat

Zero file-upload requests. Your PDF stays 100% local.

FAQ

Is my PDF uploaded to a server?
No. Your PDF is processed entirely in your browser using pdf.js. The file never leaves your device. You can verify by checking DevTools > Network — zero file-upload POST requests.
What AI model is used?
It depends on your device. WebGPU-capable browsers with enough VRAM use Llama 3.1 8B or Llama 3.2 3B locally. Otherwise, you can use your own OpenAI/Anthropic API key, or fall back to text search mode.
What's the file size limit?
50 MB maximum. Since processing is local, very large PDFs may be slower depending on your device's memory and GPU.
Does it work offline?
After the AI model is downloaded once (cached in your browser), yes. PDF processing with pdf.js always works offline.
What about the API key mode?
You can enter your own OpenAI or Anthropic API key. It's stored only in localStorage, and API calls go directly from your browser to the provider — never through our servers.
Is this really free?
Yes, completely free with no limits. Browser AI uses your GPU. API mode uses your own credits. We don't run any AI servers.
What browsers are supported?
Chrome 113+ and Edge 113+ support WebGPU for local AI. Safari and Firefox work with API key mode or text search. PDF processing works in all modern browsers.