Skip to main content
Welcome to the developer documentation for the Mixfont API! Mixfont is a typographic intelligence company. Our initial model, Lens, is the world’s first neural net trained to classify open-source fonts. Lens is the most accurate model in the world for recognizing fonts from images without manual intervention or character bounding boxes. Lens takes any input image with text in it, and maps the largest detected word area in the image to the closest Google Font in appearance. Lens can map fonts to a pretrained set of over 1000 fonts in a variety of styles. Example of the Lens model Lens is also ideal for finding an open source substitute for the any font. Because all fonts suggested by this model are open-source and released under either the OFL (Open Font License) or the Apache License, they are free to use for both personal and commercial purposes. Besides Lens, we are also working on additional font-specific AI models, including font generation, font search, and more. Stay tuned for updates!

​
Setting up the API

Before you make your first request, create an API key.
  1. Visit Mixfont and sign in.
  2. Create a key for the API.
  3. Include the key in the x-api-key header for each request.

​
API reference

Lens

Send an image URL and get the closest matching open source fonts for the largest detected word in the image.

​
Examples

You can try out the Lens model for free using the UI at https://www.mixfont.com/lens (sign in required). Here is an example output from the model: Example output from the Mixfont Lens API