Fastino Labs

Fastino Labs builds production-ready, task-specific small language models (SLMs) optimized for speed, accuracy, and edge deployment. The company's flagship GLiNER models are zero-shot SLMs for named entity recognition (NER), information extraction, and classification tasks, achieving fast inference (typically under 50ms) with a small footprint (200M parameters, under 200MB RAM) that runs on CPU hardware — including edge devices. Fastino's Pioneer fine-tuning platform allows developers to generate synthetic training datasets and fine-tune GLiNER models for domain-specific tasks, delivering 20–50% F1-score improvements over base models. With over 5 million developer users and 2,400+ GitHub stars, the company has raised a $25M seed round led by Khosla Ventures, Microsoft, and Insight Partners. Applications span PII detection and redaction, agentic AI guardrails, clinical data extraction, and ad detection.