BentoPDF

1. What is BentoPDF?

BentoPDF is a free PDF toolkit that runs right in your browser: merge files, split pages, rotate, compress, add a password, sign, turn a PDF into images, turn Word into PDF, and run OCR (turn a scanned PDF into one whose text you can select and copy). It is open source, and the project says it offers 50+ tools.

What sets it apart from the usual online PDF sites: according to the project, all processing happens on your own computer using WebAssembly (a technology that lets browsers run heavy programs), so your files are not uploaded to a server. The project states there is no sign-up, no limit on how many files you process, and no watermark. It stands in for sites such as iLovePDF or Smallpdf, which often cap daily use or ask for payment.

A good fit if:

• You handle PDFs with sensitive content (contracts, ID papers, bank statements) and do not want to send them to a random website.
• You often merge, split, compress, password-protect or OCR PDFs and do not want to install heavy software.
• You want to host a shared PDF toolbox for a team or family (Docker instructions exist).

Not a fit if:

• You need to translate whole documents: BentoPDF has no translation tool. See the FAQ below for an alternative.
• Your computer or browser is very old. Heavy jobs (OCR, Word or Excel to PDF) run on your machine and need enough power and RAM.
• You need professional-grade PDF editing like paid desktop suites. BentoPDF has an editor, but try it on your own file before relying on it for anything important.

2. What you need

The simplest way needs nothing to prepare: a modern browser (Chrome, Edge, Firefox, Safari) and an internet connection. No account, no API key, no cost.

You still need internet on first use: per the README, the heavy processing modules (such as PyMuPDF and Ghostscript) are loaded at run time from the jsDelivr CDN (a public file-delivery network for web libraries). Your PDF files themselves go nowhere.

Only if you want to host it yourself (section 3, Way B), you also need:

• Docker (software that runs apps in ready-made "boxes"). Download from https://www.docker.com/ (Docker Desktop for Windows and macOS).
• Or Node.js (a JavaScript runtime, used here to serve a static folder) from https://nodejs.org/. The README says Node.js 18 or newer.

About the license: BentoPDF is dual-licensed. The AGPL-3.0 version is free for open-source use. A one-time $79 commercial license is for people who want to embed it in closed-source products. If you only use it yourself, you can ignore this.

3. Step-by-step setup

Way A: Use it on the web (recommended for beginners)

Step 1: Open the site

Go to https://bentopdf.com in your browser. The home page lists tools in groups: organize pages, edit, convert to PDF, convert from PDF, secure and optimize.

You should see a grid of tool tiles, each named like Merge PDFs, Split PDFs or Compress PDF.

Step 2: Switch language if you like

The README lists the interface as available in several languages, including Vietnamese. Look for the language selector on the page. English works as is.

Step 3: Pick a tool and drop a file

Click the tool you need, then drag a PDF into the box or click to choose a file. For example with Merge PDFs: pick the files in the order you want, click the merge button, then download the result.

Seeing the result file land in your Downloads folder means it worked. Files are not sent to a server while processing, so large files may take a while depending on your computer.

Way B: Run it on your own computer with Docker

Use this if you want a private copy, run it on an internal network, or not depend on the public site.

Step 1: Install Docker

Install Docker Desktop from https://www.docker.com/ and open it. Wait until it reports that Docker is ready. On Windows it may ask you to enable WSL 2 and restart; follow the installer.

Step 2: Run BentoPDF

Open Terminal (macOS/Linux) or PowerShell (Windows) and run the command from the README:

docker run -p 3000:8080 ghcr.io/alam00000/bentopdf-simple:latest

The first run downloads the image, which takes a few minutes. While it runs, the window just sits there. That is normal.

Note: The bentopdf-simple image is the self-hosted build: every PDF tool is included, only the marketing parts of bentopdf.com (hero, FAQ, footer) are left out. The README says it is not a cut-down version.

Step 3: Open the page

Open your browser and go to http://localhost:3000. Seeing the BentoPDF interface means it works. To stop, go back to the terminal window and press Ctrl+C.

Way C: Download a prebuilt package (less common)

Go to https://github.com/alam00000/bentopdf/releases, download the dist-<version>.zip file, unzip it and serve the folder with a static web server. If Node.js is installed, open a terminal inside the unzipped folder and run:

npx http-server -c-1

Then visit http://localhost:8080/. This route is for people comfortable with the command line.

Note: The tool that converts Office files (Word, Excel, PowerPoint) to PDF uses LibreOffice WASM and needs a secure page address. The README says http://localhost works, while LAN addresses like http://192.168.x.x usually need HTTPS for those tools to run.

4. First use

Try one concrete job: shrink a PDF that is too heavy to email.

1. Open the site (Way A or B) and choose Compress PDF.
2. Drag your PDF into the box.
3. Pick a compression level if offered, then start.
4. Download the result and compare file sizes.

Another worthwhile first try is OCR PDF, to make a scan selectable and copyable:

1. Choose OCR PDF and drop in a scanned file.
2. Pick the language(s) in the document from the searchable list. The project docs say it uses Tesseract and supports many languages, several at once. The language list in the project's packaging script includes Vietnamese, so pick it for Vietnamese documents.
3. Click Start OCR and wait for the progress bar.
4. Download the searchable PDF, or a plain .txt file.

OCR quality depends on how clear the scan is. If many words come out wrong, try higher resolution (High or Ultra) or turn on Binarize (convert to black and white) for faded scans, as the project docs suggest.

5. Common problems and fixes

Symptom	Likely cause	Fix
A tool spins forever with no result	File too large for your RAM, or the tab hit its memory limit	Try a smaller file, close other tabs, try Chrome or Edge
A tool fails with a network error	Your network blocks jsDelivr, or you are offline (modules are loaded at run time)	Switch network. For a closed network, see the "air-gapped" section of the README
Word/Excel to PDF will not run when opened over the LAN	It needs a secure address (HTTPS); http://192.168... is not enough	Open it via http://localhost, or put HTTPS in front
docker: command not found	Docker is not installed or Docker Desktop is not running	Install Docker, open Docker Desktop, reopen the terminal and retry
docker run says port 3000 is already in use	Another program uses port 3000	Change the first number in -p, for example -p 3100:8080, then visit http://localhost:3100
OCR text is mostly wrong	Wrong language chosen, or a blurry scan	Pick the right language, raise the resolution, try Binarize

6. Uninstall / update

• Way A (web): nothing to remove. The site is always the latest version.
• Way B (Docker): press Ctrl+C to stop. To remove it completely, open Docker Desktop, delete the BentoPDF container under Containers and the bentopdf-simple image under Images. To update, run docker pull ghcr.io/alam00000/bentopdf-simple:latest and then run the docker run command again.
• Way C: delete the unzipped folder. To update, download the new dist package from the Releases page.

7. FAQ

Does it cost money? The project says it is free, unlimited and needs no sign-up. Only embedding it in a closed-source product needs the $79 commercial license.

Are my files uploaded? According to the project docs, no: processing happens in your browser. The processing modules are still downloaded from a CDN at run time, so you need internet on first use.

Is there Vietnamese? The README lists a Vietnamese interface, and the OCR language list in the packaging script includes Vietnamese.

What if I want to translate a whole PDF? BentoPDF does not translate. See PDFMathTranslate on githot: https://www.githot.io.vn/en/repo/pdfmathtranslate

Will it run on a weak computer? Light jobs like merge, split and rotate are usually fine. OCR and Office conversion are heavier because they run on your machine. The docs give no minimum RAM, so try a small file first.