A 164-megabyte speech model transcribes an hour of audio in about twenty seconds on a laptop
Fermion Research built it by compressing Nvidia's 2.5-gigabyte Parakeet model until each encoder weight holds one of five learned levels, at about 2.1 bits.
Model
Phonon-2
Phonon-2 is an artificial intelligence model developed by Fermion Research. This page lists what is publicly known about the model and every benchmark result Aivio News has recorded for it, each with its own source and date.
- Developer
- Fermion Research
- Model
- Phonon
- Released
- September 30, 2026
- Modalities
- Audio, Model
- Status
- Active
- Source
- fermionresearch.com
Benchmarks
All storiesNo verified results recorded yet.