Magnitude runs open AI models up to 2x faster by tuning to your hardware

Original: Launch HN: Magnitude (YC S25) – Self-optimizing inference engine for agents

Source: github.com ↗