Kokoro v1
Kokoro
Key facts · verified Oct 13, 2026
Kokoro v1 is a Kokoro audio generation model released on 27 Jan 2025. Pricing: Open weights — free to self-host. Source: huggingface.co.
Kokoro, an 82-million-parameter open-weight text-to-speech model under Apache 2.0. Version 1.0 was published on 27 January 2025 with 54 voices across 8 languages, and its authors note it has been deployed in commercial APIs.
Pricing & specs checkedSource: vendor documentation · checked Oct 13, 2026LMArena: 1200 Elo (last listed Aug 22, 2026) · source · retrieved Aug 22, 2026
Key Features
Audio Generation Text-to-Speech
Full Specifications
Open weights — free to self-host | |
|---|---|
| Identity | |
| Developer | Kokoro |
| Released | Jan 2025 |
| Licence | Apache 2.0 |
| Self-hostable | Yes — 82M parameters |
| Capability | |
| Languages | 8 languages, 54 voices (v1.0) |
| Rights | |
| Commercial licence | Yes — Apache 2.0 |
| Measured quality | |
| LMArena Elo | 1200 (checked Aug 2026) |
| LMArena rank | Rank 10 |