User Manual

1. Common

• The home screen has 4 categories: "Create & edit PDF / Convert & compose PDF / OCR & AI-OCR / Settings". Tap a category to see its tools.
• Input files are read-only; the original is never modified. Results are always saved as new files.
• Choose the output location in each screen's "Output folder". If unset, files are saved to "Download/PDFToolkit" on the device.

2. Edit PDF (add text/images, edit existing text)

1. Open a PDF with "Choose PDF", or make an empty A4 with "Start from blank".
2. The preview stays fixed as a canvas; use the toolbox at the bottom to move between pages and pick each action.
3. Add text: pick "Text" from the toolbox, type into the floating window (line breaks allowed), set size / color / font / bold / italic / underline / rotation / URL, then "Add". It is placed on the preview; drag to move it.
4. Add an image: choose one under "Image" and it is placed on the preview. Drag to move it; you can also scale and rotate it.
5. Edit existing text: tap text on the preview, or select it from the "Layers" list. Enter "Replacement text", or choose "Delete the original text". You can also change size / color / styling.
   • If the font and character set match, it is replaced in place.
   • For characters that can't be shown, moves, or size/color/style/font changes, the whole run is redrawn, keeping the original size and color (using the chosen font).
6. Layers: added/edited items and recognized objects on the page are listed under "Layers". Tap a row to select it, or × to remove it.
7. Tapping "Apply" in each menu, or in the Layers list, bakes the current edits into a temporary PDF and refreshes the preview with the real appearance.
8. "Apply and save" outputs the final PDF.

* Adding/editing text needs an embedded font. Choose from Noto Sans JP / Noto Serif JP / M PLUS Rounded 1c / Zen Kaku Gothic New / Klee One (all SIL OFL), per text run. Each font downloads once, then works offline; get missing ones from the editor or Settings.

3. Convert & compose PDF

• Split (extract pages): open a PDF, select the pages to extract, then choose "combine into one" or "one per page" and export.
• Merge: choose several PDFs, arrange the order, and combine into one.
• Reorder: drag thumbnails to change the page order and save as a new PDF.
• PDF to images: convert each page to PNG / JPEG. You can set the DPI (resolution).
• Images to PDF: choose several images, set the order, and combine into one PDF.

4. OCR / AI-OCR

• OCR / text extraction: extract text from PDFs/images. Embedded text (already in the file) and OCR (image recognition) are distinguished, and can also be exported as JSON. Engines: Tesseract / PaddleOCR / local LLM Vision, with Japanese, English, Simplified Chinese, and Korean recognition. Large documents can run in the background.
• PDF summary: summarize the whole file or per page with an LLM. Choose "OCR→LLM" or "Vision (page images sent directly to the LLM)".
• AI Prompt: enter an arbitrary instruction (prompt) and run it over the whole PDF or per page. Choose "OCR→LLM" or "Vision".

5. Settings

• OCR settings and models: download Tesseract / PaddleOCR language models (Japanese, English, Chinese, Korean) and the editing fonts, and set the LLM (Ollama / OpenAI-compatible) connection URL, model, and connection test. Order is Tesseract → PaddleOCR → LLM.
• Launch LLM API: the "Launch LLM API" button starts the companion app (LLM Tester) API server, and a green/amber status tile shows the /api/tags reachability.
• User manual / Privacy policy / Licenses: these documents (you can copy the body on each screen).

6. About offline use

• PDF and image processing run entirely on the device. The only network use is "the first download of OCR models / fonts" and "sending to the configured server when using an LLM" (the default LLM endpoint is 127.0.0.1 on the device).
• If you configure an external LLM server, page images and extracted text are sent there. The destination depends on your own settings.