Kokoro
FreeAn 82M parameter TTS model by Hexgrad.
CompactOpen-SourceEfficient
Overview
Kokoro is a surprisingly capable 82M parameter model that punches well above its weight, delivering natural speech at a fraction of the compute of larger systems. Its tiny footprint makes it ideal for real-time and edge use cases. Released under a permissive Apache-2.0 license, it has quickly become a community favourite.
What makes Kokoro special?
- Focus: Compact high-quality synthesis.
- Availability: Free & Open-Source (Apache-2.0).
- Use cases: Edge devices, real-time apps, and cost-sensitive deployments.