Working with non-code files

GitHub supports rendering and diffing in a number of non-code file formats.

Rendering and diffing images

GitHub can display several common image formats, including PNG, JPG, GIF, PSD, and SVG. In addition to simply displaying them, there are several ways to compare differences between versions of those image formats.'

Note: If you are using the Firefox browser, SVGs on GitHub may not render.

Viewing images

You can directly browse and view images in your GitHub repository:

inline image

SVGs don't currently support inline scripting or animation.

Viewing differences

You can visually compare images in three different modes: 2-up, swipe, and onion skin.


2-up is the default mode; it gives you a quick glimpse of both images. In addition, if the image has changed size between versions, the actual dimension change is displayed. This should make it very apparent when things are resized, such as when assets are upgraded to higher resolutions.



Swipe lets you view portions of your image side by side. Not sure if colors shifted between different versions? Drag the swipe slider over the area in question and compare the pixels for yourself.


Onion skin

Onion Skin really comes in handy when elements move around by small, hard to notice amounts. Did an icon shift two pixels to the left? Drag the opacity slider back a bit and notice if things move around.

Onion skin

3D File Viewer

GitHub can host and render 3D files with the .stl extension.

When looking directly at an STL file on GitHub you can:

  • Click and drag to spin the model.
  • Right click and drag to translate the view.
  • Scroll to zoom in and out.
  • Click the different view modes to change the view.


When looking at a commit or set of changes which includes an STL file, you'll be able to see a before and after diff of the file.

By default, you'll get a view where everything unchanged is in wireframe. Additions are colored in green, and removed parts are colored in red.


You can also select the Revision Slider option, which lets you use a slider at the top of the file to transition between the current and previous revisions.

Fixing slow performance

If you see this icon in the corner of the viewer, then the WebGL technology is not available on your browser:

WebGL pop error

WebGL is necessary to take advantage of your computer's hardware to its fullest. We recommend you try browsers like Chrome or Firefox, which ship with WebGL enabled.

Error: "Unable to display"

If your model is invalid, GitHub may not be able to display the file. In addition, files that are larger than 10 MB are too big for GitHub to display.

Embedding your model elsewhere

To display your 3D file elsewhere on the internet, modify this template and place it on any HTML page that supports JavaScript:

<script src="<username>/<repo>/<ref>/<path_to_file>"></script>

For example, if your model's URL is, your embed code would be:

<script src=""></script>

By default, the embedded renderer is 420 pixels wide by 620 pixels high, but you can customize the output by passing height and width variables as parameters at the end of the URL, such as ?height=300&width=500.

Note: ref can be a branch or the hash to an individual commit (like 2391ae).

Rendering CSV and TSV data

GitHub supports rendering tabular data in the form of .csv (comma-separated) and .tsv (tab-separated) files.

Rendered CSV sample

When viewed, any .csv or .tsv file committed to a GitHub repository automatically renders as an interactive table, complete with headers and row numbering. By default, we'll always assume the first row is your header row.

You can link to a particular row by clicking the row number, or select multiple rows by holding down the shift key. Just copy the URL and send it to a friend.

Searching data

If you want to find a certain value in your dataset, you can start typing in the search bar directly above the file. The rows will filter automatically:

Searching for values

Handling errors

Occasionally, you may discover that your CSV or TSV file isn't rendering. In those instances, an error box appears at the bottom of your raw text, suggesting what the error may be.

CSV render error message

Common errors include:

  • Mismatched column counts. You must have the same number of separators in each row, even if the cell is blank
  • Exceeding the file size. Our rendering only works for files up to 512KB. Anything bigger than that slows down the browser.

Rendering PDF documents

GitHub supports rendering of PDF documents.

Rendered PDF Document

Currently, links within PDFs are ignored.

Rendering differences in prose documents

Commits and pull requests that include prose documents have the ability to represent those documents with source and rendered views.

The source view shows the raw text that has been typed, while the rendered view shows how that text would look once it's rendered on GitHub. For example, this might be the difference between showing **bold** in Markdown, and bold in the rendered view.

Prose rendering is supported for rendered documents supported by github/markup:

  • Markdown
  • AsciiDoc
  • Textile
  • ReStructuredText
  • Rdoc
  • Org
  • Creole
  • MediaWiki
  • Pod

Paper icon to view rendered prose document

You can click to see the changes made to the document as part of a commit.

Rendered Prose changes

Visualizing attribute changes

We provide a tooltip describing changes to attributes that, unlike words, would not otherwise be visible in the rendered document. For example, if a link URL changes from one website to another, we'd show a tooltip like this:

Rendered Prose attribute changes

Commenting on changes

Commit comments can only be added to files within the source view, on a line-by-line basis.

Linking to headers

As with other rendered prose documents, hovering over a header in your document creates a link icon. You can link readers of your rendered prose diff to specific sections.

Viewing complex diffs

Some pull requests involve a large number of changes with large, complex documents. When the changes take too long to analyze, GitHub can't always produce a rendered view of the changes. If this happens, you'll see an error message when you click the rendered button.

Message when view can't be rendered

You can still use the source view to analyze and comment on changes.

Viewing HTML elements

We don't directly support rendered views of commits to HTML documents. Some formats, such as Markdown, let you embed arbitrary HTML in a document. When these documents are shown on GitHub, some of that embedded HTML can be shown in a preview, while some (like an embedded YouTube video) cannot.

In general, rendered views of changes to a document containing embedded HTML will show changes to the elements that are supported in GitHub's view of the document. Changes to documents containing embedded HTML should always be reviewed in both the rendered and source views for completeness.

Mapping geoJSON files on GitHub

GitHub supports rendering geoJSON and topoJSON map files within GitHub repositories. Simply commit the file as you would normally using a .geojson or .topojson extension. Files with a .json extension are also supported, but only if type is set to FeatureCollection, GeometryCollection, or topology. Then, navigate to the path of the geoJSON file on

When you click the paper icon on the right, you'll also see the changes made to that file as part of a commit.

Source Render toggle screenshot

Geometry Types

Maps on GitHub use Leaflet.js and support all the geometry types outlined in the geoJSON spec (Point, LineString, Polygon, MultiPoint, MultiLineString, MultiPolygon, and GeometryCollection). TopoJSON files should be type "Topology" and adhere to the topoJSON spec.

Styling features

You can customize the way features are displayed, such as specifying a particular color or adding a descriptive icon, by passing additional metadata within the geoJSON object's properties. The options are:

  • marker-size - small, medium, or large
  • marker-color - valid RGB hex color
  • marker-symbol - an icon ID from the Maki project or a single alphanumeric character (a-z or 0-9).
  • stroke - color of a polygon edge or line (RGB)
  • stroke-opacity - opacity of a polygon edge or line (0.0 - 1.0)
  • stroke-width - width of a polygon edge or line
  • fill - the color of the interior of a polygon (GRB)
  • fill-opacity - the opacity of the interior of a polygon (0.0-1.0)

See version 1.1.0 of the open simplestyle spec for more information.

Embedding your map elsewhere

Want to make your geoJSON map available someplace other than GitHub? Simply modify this template, and place it in any HTML page that supports javascript (e.g., GitHub Pages):

<script src="<username>/<repo>/<ref>/<path_to_file>"></script>

For example, if your map's URL is, your embed code would be:

<script src=""></script>

By default, the embedded map 420px x 620px, but you can customize the output by passing height and width variables as parameters at the end, such as ?height=300&width=500.

Note: ref can be a branch or the hash to an individual commit (like 2391ae).


If your map contains a large number of markers (roughly over 750), GitHub will automatically cluster nearby markers at higher zoom levels. Simply click the cluster or zoom in to see individual markers.

Something's up with the underlying map

The underlying map data (street names, roads, etc.) are driven by OpenStreetMap, a collaborative project to create a free editable map of the world. If you notice something's not quite right, since it's open source, simply sign up and submit a fix.


If you're having trouble rendering geoJSON files, ensure you have a valid geoJSON file by running it through a geoJSON linter. If your points aren't appearing where you'd expect (e.g., in the middle of the ocean), it's likely that the data is in a projection which is currently unsupported. Currently, GitHub only supports the urn:ogc:def:crs:OGC:1.3:CRS84 projection.

Additionally, if your .geojson file is especially large (over 10 MB), it is not possible to render within the browser. If that's the case, you'll generally see a message that looks something like this:

Large file

It may still be possible to render the data by converting the .geojson file to TopoJSON, a compression format that, in some cases, can reduce filesize by up to 80%. Of course, you can always break the file into smaller chunks (such as by state or by year), and store the data as multiple files within the repository.

Additional Resources

Working with Jupyter Notebook files on GitHub

When you add Jupyter Notebook or IPython Notebook files with a .ipynb extension on GitHub, they will render as static HTML files in your repository.

The interactive features of the notebook, such as custom JavaScript plots, will not work in your repository on GitHub. For an example, see Linking and Interactions.ipynb.

To view your Jupyter notebook with JavaScript content rendered or to share your notebook files with others you can use nbviewer. For an example, see Linking and Interactions.ipynb rendered on nbviewer.

To view a fully interactive version of your Jupyter Notebook, you can set up a notebook server locally. For more information, see Jupyter's official documentation.


If you're having trouble rendering Jupyter Notebook files in static HTML, you can convert the file locally on the command line by using the nbconvert command:

$ jupyter nbconvert --to html NOTEBOOK-NAME.ipynb

Further reading

Esse documento ajudou você?Política de Privacidade

Ajude-nos a tornar esses documentos ótimos!

Todos os documentos do GitHub são de código aberto. Você percebeu que algo que está errado ou não está claro? Envie um pull request.

Faça uma contribuição

Ou, aprenda como contribuir.