Fastino Releases GLiNER2.5: A Boundary-Prediction Architecture That Removes Span Enumeration From Information Extraction

Fastino has released GLiNER2.5, an information extraction architecture that replaces span enumeration with boundary prediction to reduce computational costs. By decoupling entity width from processing requirements, the update allows for more efficient performance across its three CPU-runnable model sizes. The release also introduces support for joint entity-relation decoding and constrained classification, broadening the tool's utility for structured data extraction tasks.