TypeSafe's Jev Model Accelerates Building of Content-Detection Tools Like X Slop Detectors

TypeSafe AI's newly launched Jev model is democratizing the creation of content-screening tools, making it practical to rapidly build systems that detect low-quality posts and AI-generated "slop" on social platforms. The startup, founded by Diego Almeida—a co-creator of ChatGPT and developer of RLHF at OpenAI—released Jev into early access on September 15, 2026, backed by a $40 million seed round led by DCVC. Jev represents a new class of models called "System One" architectures: instead of generating text autoregressively like traditional language models, it returns typed, probabilistic decisions in parallel within 70 to 500 milliseconds. The model costs $42 per billion input tokens with free output, and Jev claims 40- to 200-fold speedups over frontier language models on classification and decision tasks.