The Neural Amp Modeler Plugin is the free, open-source audio plugin that plays Neural Amp Modeler (NAM) captures: neural-network models of real guitar amplifiers and pedals. Created by Steven Atkinson (sdatkinson) under the MIT license, it has about 3,050 stars on GitHub and is actively maintained in 2026. NAM has become one of the most popular open ecosystems in guitar technology, with thousands of community-shared amp captures available for it.
The plugin loads .nam model files (captures trained from recordings of physical amps, drive pedals and full rigs) and runs them in real time inside your DAW as a VST3, AU or AAX plugin, or as a standalone app. A built-in signal chain provides a noise gate, input/output controls, a tone stack (EQ) and an impulse-response cabinet loader, so a single capture can be shaped into a complete guitar tone. Because inference is a small neural network running on CPU, latency stays low enough for live playing and tracking.
The plugin is one half of an open ecosystem: its sibling project, the Neural Amp Modeler trainer, lets anyone create a capture by recording their own amp through a re-amping process and training a model on the result. Captures are freely traded by the community, so the plugin effectively gives you a growing library of boutique and classic amplifiers at no cost. Under the hood the plugin is built on the iPlug2 framework, and its efficient CPU inference means it can run many instances in a session alongside other plugins.
The plugin is completely free and open source under the MIT license, and the wider NAM ecosystem of shared captures is community-driven and free.
Guitarists, producers and home-studio musicians who want realistic amp tones in the box without buying commercial modelling software or hardware.