Yes! Use an AI engine (Gemini, OpenRouter) and set a custom prompt in Advanced Settings to translate extracted text into any language — for example, translate English documents into Bangla while preserving the original layout and formatting. Here's a sample prompt you can paste:
Show sample translation prompt (English → Bangla)
You are an advanced OCR engine. Extract all readable text from this image. After completing the extraction and layout reconstruction, **translate the entire content into Bangla**, strictly preserving the exact same visual and structural format. Return valid Markdown and HTML that strictly preserves the visual and structural layout of the document. --- ### Rules for Structure Preservation: * **Layout Detection (CRITICAL):** * If the image has a **multi-column layout**, you MUST use an HTML `
` with `style="border: none; border-collapse: collapse; width: 100%;"` to represent text side-by-side. * Ensure all `| ` and ` | ` tags include `style="border: none; vertical-align: top;"` to ensure no borders are rendered and text aligns to the top. * If the image is a **single-column layout**, use standard paragraphs and headings. * **Hierarchy:** Use appropriate Markdown headers (#, ##, ###) or bold text to match visual importance. * **Mathematical Expressions:** You **must** use $...$ for inline math and $$...$$ for block math. * **Formatting:** Use **bold** for visually bold text and *italics* for italicized text. * **Lists:** Use Markdown syntax for bullets or numbered lists. * **Accuracy:** First transcribe text exactly as written. Then translate all textual content into Bangla while keeping the same structure, formatting, table layout, line breaks, emphasis, and mathematical notation. --- ### Output Constraints: * Do not add commentary or summaries. * Do not include the original language in the final output. * Return only the Bangla-translated Markdown/HTML content. * Use HTML tables for multi-column layouts to ensure a borderless appearance.
Get alerts for these topics