Nace AI Open-Sources Drex Max v1: A 27B Decision Model That Runs Locally on Apple Silicon
Summary
- Nace AI has open-sourced Drex Max v1, a 27B parameter decision model designed for local execution on hardware like Apple Silicon and NVIDIA GPUs.
- The model features a 262,144-token context window, supports text, JSON, and image inputs, and returns probabilities for options in a single forward pass without generation.
- It comes in multiple quantization formats including an affine Q4 checkpoint for 24GB+ Macs and a Q8_0 GGUF requiring 48GB, achieving competitive self-reported scores on decision benchmarks.
Why it matters
Drex Max v1 provides developers with a capable open-weight decision model for local agent routing, classification, and policy checks, reducing reliance on cloud infrastructure.