Musubi Releases PolicyLM-1.7B, an Open-Weights Moderation Model for Real-Time Content Screening

Musubi, an AI-powered content moderation startup, released PolicyLM-1.7B, a lightweight decision model designed for real-time moderation of text across social platforms, messaging services, and user-generated content feeds. The model is distributed with open weights under the Apache 2.0 license, available on Hugging Face and deployable on-premises, offering platforms an alternative to closed proprietary moderation APIs.