Forge is Mixfont's font generation model, built around taking an image reference and turning that direction into an original font system.
Forge is designed to start with the same kind of visual reference teams already collect during concepting: packaging crops, scans, posters, screenshots, and lettering samples with a strong point of view.
Instead of stopping at recognition, Forge is being built to turn those references into original font assets. The model is still coming soon, but the goal is clear: move from inspiration to usable type without leaving Mixfont.
Forge is built around the idea that a strong image can communicate enough about a lettering style to guide font creation. A screenshot, poster crop, packaging detail, or scan can all serve as direction when the typography clearly shows mood, contrast, rhythm, and proportion.
That reference is meant to describe intent rather than provide a source file. The model uses the visual cues in the image to infer how the broader type system should feel.
Rather than finding the nearest existing family, Forge is being built to synthesize new glyph shapes that stay coherent across a full font. The goal is a usable type asset, not a one-off lettering reconstruction.
That makes Forge a different workflow from Lens. Lens helps you identify what already exists, while Forge is aimed at extrapolating from a visual direction and turning it into something original.
Forge is still in training, so Mixfont currently treats it as a coming-soon workflow while the generation controls, review loop, and product details are refined.
As the model rolls out, Forge is expected to live inside the same Mixfont platform and API ecosystem as Lens so teams can move from inspiration to generated font files in a familiar workflow.
Forge is Mixfont’s font generation model. It is being built to take an image reference and turn that visual direction into an original font rather than simply identify the closest existing family.
The clearest inputs are images where the lettering style is easy to read and visually distinctive. References with obvious contrast, proportions, terminals, and rhythm give the model more useful design signals to work from.
Lens is a recognition model that returns ranked open-source font matches from an image. Forge is a generation model that aims to use the image as creative direction for producing a new typeface.
Forge is intended to use the reference as stylistic guidance, not as a way to retrieve or redistribute an existing font file. The product direction is original font generation informed by the image you upload.
The goal is a usable generated font file that can be reviewed and deployed inside the Mixfont workflow. Mixfont has not published the full release spec yet, so the exact controls and export details are still evolving.
Not yet. As of March 13, 2026, Forge is still marked coming soon and is not broadly available in Mixfont.
Mixfont’s current positioning is that Forge will be available through the platform and API as it launches. Final release timing and access details have not been published yet.
Coming soon
Forge is still in training. Sign in to explore Mixfont now and be ready when image-to-font generation opens up.
