Musubi launches PolicyLM-1.7B for real-time content moderation

Startup Musubi has released PolicyLM-1.7B, an open-weight decision model that applies plain-English content policies to messages in under 50 milliseconds. The model needs no retraining when policies change and matches the cost and speed of classifiers used by major platforms.

• 1.7B-parameter model released with open weights
• Applies policies to messages in under 50 ms
• Policies written in plain English, no retraining needed
• Cost and speed comparable to social platform classifiers