Self-Supervised Flow Matching for Scalable Multi-Modal Synthesis

Black Forest Labs · Project update · March 3, 2026

The company released a research report detailing a self-supervised flow matching framework. This architecture enables the scalable synthesis of multiple modalities, including image, video, and audio, within a single system.

Read the report

AntSpace’s summary of Black Forest Labs’s own announcement: read the original.

More updates from Black Forest Labs