Show HN: Tiny model for fast typed decisions on CPU (Gutsy)

Category: ai-ml

Tags: decision-making, local-inference, calibration, cpu, model-serving

Score: 7.5/10 (Innovation: 7, Technical: 7, Documentation: 9, Utility: 7)

Gutsy provides a small, locally-runnable 0.8B model for structured decision-making, returning calibrated probabilities for yes/no, multiple-choice, and scoring questions via a simple HTTP API. It stands out for its focus on calibrated outputs, deterministic inference, and CPU-friendly operation, making it a practical alternative to larger models for classification and routing tasks. The project includes thorough documentation, benchmarks, and honest limitations.

Target audience: backend developers, data scientists, AI engineers

Repository: https://github.com/kouhxp/gutsy · HTML · Apache-2.0 · 94 stars

View on Hacker News