# Image to Markdown converter with local OCR

> Open a PNG, JPG, WebP, or BMP image to get a Markdown document that embeds it, lists its details, and, if you turn on OCR, includes the English text found in the image. OCR runs in your browser.

Source: https://markdownlens.ayushdev.com/image-to-markdown · Updated 2026-11-12 · Markdown Lens, free and open source: https://github.com/ayushhagarwal/markdown-lens

The original image is kept in the local document workspace. OCR is opt-in, cancelable, and does not send pixels to a hosted service.

Open the converter: https://markdownlens.ayushdev.com/image-to-markdown (drop a PNG, JPG, WebP, or BMP file on the page; it is processed in your browser and not uploaded).

## Steps

1. Choose a supported image.
2. Decide whether to run local English OCR.
3. Review recognized text and export the image bundle.

## Example: the Markdown an image produces

Input: `whiteboard.png` (Image with OCR turned on)

- `1600 × 900 PNG photo of a whiteboard`

Markdown output:

````markdown
# whiteboard

![whiteboard](assets/whiteboard.png)

## Image details

- Dimensions: 1600 × 900
- Type: image/png
- Size: 412.3 KB

## Extracted text

```text
(text recognized in the image appears here)
```
````

The structure is exactly what the converter writes; the size and the recognized text depend on your image. Without OCR, the Extracted text section is left out.

## What the converter preserves

- Original image as a local document asset
- Image dimensions, type, and source size
- Optional locally recognized English text
- Portable Markdown-plus-assets bundle export

## Limitations

- OCR accuracy depends on resolution, contrast, orientation, and typography.
- The bundled OCR language is English in this release.
- OCR extracts text but does not infer chart meaning or visual semantics.
- Scanned PDFs are a different workflow and still need a dedicated OCR step before PDF conversion.

## How to convert an image to Markdown

1. Drop a PNG, JPG, WebP, or BMP file on this page.
2. Choose whether to run English OCR. The OCR engine loads from this site and runs on your device.
3. Correct any misread words in the extracted text.
4. Export a bundle so the image file travels with the Markdown that links to it.

## Frequently asked questions

### Is OCR cloud-based?

No. The OCR worker and English model run locally after they are loaded from the same site origin.

### Does it describe charts?

No. This release extracts visible text and image metadata rather than interpreting visual meaning.

### Can it extract images from a PDF?

No. The PDF converter extracts text only. To get Markdown with local image links, save the PDF pages or figures as images and import those here, or convert the original Word or PowerPoint file, which keeps embedded images as assets.

## Related

- [PDF to Markdown converter](https://markdownlens.ayushdev.com/pdf-to-markdown): Convert text-based PDFs locally. Scanned PDFs still need OCR first.
- [Word to Markdown converter](https://markdownlens.ayushdev.com/word-to-markdown): Convert .docx files and keep embedded images as local assets.
