---
title: "Convert a PDF to an image in C#"
url: https://recursiveai.org/oneimage/cookbook/pdf-to-image/
description: "Convert PDF pages to PNG or JPG in C# with pure .NET code, no PDFium or Ghostscript: every page, one page at 300 dpi, or a thumbnail. Runnable code with OneImage.Pdf."
version: "1.10.4"
---

# Convert a PDF to an image in C#

Add the `RecursiveAI.OneImage.Pdf` package, register `PdfFormat` in the load options, and load the PDF: each page is a frame. Save `Frames[i].ToImage()` as `.png` or `.jpg`. There is no PDFium, Ghostscript or other native binary to deploy. The program below renders every page of the IRS Form W-9, a public six-page PDF.

```shell
dotnet new console -o PdfToImage && cd PdfToImage
dotnet add package RecursiveAI.OneImage.Pdf --version 1.10.4
curl -LO https://www.irs.gov/pub/irs-pdf/fw9.pdf
```

OneImage needs the .NET 10 SDK. Put this in `Program.cs` and run `dotnet run`:

```csharp
using OneImage;
using OneImage.Pdf;

// PDF is an add-on format: register it, drawn at 150 dpi.
var options = new LoadOptions
{
    Configuration = new ImageConfiguration()
        .Add(new PdfFormat(new PdfRenderOptions { Dpi = 150 })),
};

Image pdf = Image.Load("fw9.pdf", options);   // reads the structure, draws nothing
Console.WriteLine($"{pdf.FrameCount} pages");

for (int i = 0; i < pdf.FrameCount; i++)
{
    Image page = pdf.Frames[i].ToImage();
    page.Save($"page-{i + 1}.png");
    Console.WriteLine($"page-{i + 1}.png: {page.Width}x{page.Height}");
}
```

```text
6 pages
page-1.png: 1275x1650
page-2.png: 1275x1650
page-3.png: 1275x1650
page-4.png: 1275x1650
page-5.png: 1275x1650
page-6.png: 1275x1650
```

A US Letter page is 612×792 points, so at 150 dpi it is 1275×1650 pixels: points × dpi ÷ 72. Pages are drawn on opaque white, as viewers show them. The output format follows the extension, so `.jpg` and `.webp` work the same way.

## One page at 300 dpi, and a thumbnail

To render only some pages, set `FirstPage` (zero-based) and `PageCount`. A resize right after the load draws the page at the final size instead of drawing it large and shrinking it:

```csharp
using OneImage;
using OneImage.Pdf;

// Page 1 only, at 300 dpi, as JPEG.
var page1 = new LoadOptions
{
    Configuration = new ImageConfiguration().Add(new PdfFormat(
        new PdfRenderOptions { Dpi = 300, FirstPage = 0, PageCount = 1 })),
};
Image.Load("fw9.pdf", page1).Save("page-1-300dpi.jpg", new SaveOptions
{
    EncoderOptions = new JpegEncoderOptions { Quality = 85 },
});

// A 300 px wide WebP thumbnail of the first page.
Image.Thumbnail("fw9.pdf", 300, loadOptions: page1).Save("cover.webp");

foreach (string file in new[] { "page-1-300dpi.jpg", "cover.webp" })
{
    ImageInfo info = Image.Identify(file);
    long kb = new FileInfo(file).Length / 1024;
    Console.WriteLine($"{file}: {info.Width}x{info.Height}, {kb} KB");
}
```

```text
page-1-300dpi.jpg: 2550x3300, 1186 KB
cover.webp: 300x388, 21 KB
```

| `Dpi` | US Letter in pixels | Typical use |
|---|---|---|
| 72 | 612×792 | Previews |
| 150 (default) | 1275×1650 | Readable page images |
| 300 | 2550×3300 | OCR and print-quality archives |

`PdfRenderOptions` also sets the page box, whether annotations are drawn, a transparent `Background`, and the `Password` of an encrypted file.

## Limits

- **It renders pixels only.** To get the text out of a PDF, see [Extract text from a PDF](https://recursiveai.org/onepdf/cookbook/extract-text/) in OnePDF.
- **Untrusted files are bounded.** File size, page size, image pixels and nesting depth are checked from the file before memory is allocated; a file over a limit throws `ImageLimitException`. The [limits table](https://recursiveai.org/oneimage/docs/pdf/#limits) has the defaults.
- **JPEG and BMP hold one page.** Saving the whole document as `.jpg` writes page 1 only; loop over `Frames`, or [save a multipage TIFF](https://recursiveai.org/oneimage/cookbook/pdf-to-tiff/).
- **.NET 10 only.** [PDFtoImage](https://www.nuget.org/packages/PDFtoImage) (MIT) wraps native PDFium and supports older frameworks.

How pages are drawn and checked is in the [OneImage.Pdf docs](https://recursiveai.org/oneimage/docs/pdf/).