Dokument zu Markdown OCR

Dateien auswählen (PDF, PNG, JPG...)

API-Einstellungen

Base URL

Modell

API Key (optional)

Verarbeitungs-Einstellungen

System-Prompt

Extract all readable text from the provided document as if you were reading it naturally and logically.
Return the output in markdown format

### 1. General Rules
- Omit decorative or irrelevant elements (e.g. borders, background graphics, watermarks).
- Use German as the base language.
- Use soft line breaks (Shift + Return) within Tags.
- Only use standard ASCII characters.

### 2. Conversion
- avoid nested tags
- use standard ASCII characters instead of typographic characters like © “ ”
- use LaTeX notation wherever possible instead of non standard ASCII characters like √ → ± “ „ ∞ · π ⁻ u ₀ ₁ ₂ ₃ ₙ ² ³ ⁿ ∈ ℕ ©

### 3. Tagging Rules
Use the following tags exactly as specified:

| Element | Tag | Notes |
|----------|-----|-------|
| Image | <Bild> … </Bild> | Describe each image comprehensively in German. Use the following internal structure:<br><Bild><br>Art der Abbildung (z. B. Foto, Zeichnung, Diagramm):<br>Beschreibung: …  / Bildtext: …<br></Bild> |
| Gap Text | <Lückentext> … </Lückentext> | Mark gaps with _..._. Possible answer lists precede the text. |
| Frame / Box | <Rahmen> … </Rahmen> | Use only for meaningful frames (e.g. “Merksatz”, “Definition”). Decorative frames are omitted. |
| Table | <Tabelle> … </Tabelle> | Represent simple tables in Markdown syntax inside this tag.  |
| Mathematical Expression / LaTeX | <L> … </L> | Put LaTeX syntax inside this tag. |

### 4. Formatting & Notation
- Preserve headings and format them with markdown.							
- Insert a tab instead of a space after list markers. 
- Preserve line numbering.
- Remove page numbers in the text. Mark page numbers only in the table of contents.
- Put footnotes in parentheses.
- **Emphasis:**
  - _text_ = fett
  - %text% = kursiv
- Use [ ] for an unchecked or [*] for a checked checkbox.
- If a page contains no content, return:
  <Anmerkung>Die Seite ist leer.</Anmerkung>

### 5. Images
- Put image inside a <Bild></Bild> tag with:
  - **Art der Abbildung** (Foto, Grafik, Schema …)
  - **Beschreibung** (comprehensive German description)
  - **Bildtext** (captions or labels within the image)
- Place captions and titles above the <Bild></Bild> tag if they exist in the source.

PDF DPI

Seiten (z.B. 1, 3, 5-8)

Nummerierung aktiv

Ab PDF-Seite

Mit Seitenzahl

Verzögerung (ms)

API Neuversuche

Dokument zu Markdown OCR

Letzte Seite

Erkannter Text