F13 Raises $5M for Vector Graphics AI

F13 has emerged from stealth with $5M to tackle a weakness in mainstream image generation: precise, editable graphics.

What happened

The startup is building foundation models that generate structured vector graphics rather than raster images.

Its target use cases include diagrams, charts, maps, presentations and technical visuals where geometry, editability and semantic structure matter more than photographic realism.

Why it matters

Diffusion-style image models are strong at aesthetics but often struggle with exact layouts, labels and geometry.

Vector-native generation could be more useful for professional design and developer workflows because the output remains editable rather than collapsing everything into pixels.

The bigger picture

Generative design is fragmenting into specialised model types. The next wave may include systems optimised for CAD, diagrams, UI, documents and other structured outputs where accuracy matters more than visual style.