Skip to content
local-ai

Kokoro 82M

Hexgrad · Text to speech · 82M · Released 9 January 2025

Commercial use permitted TextAudio Open weights Runs on CPU Apple Silicon

Kokoro is proof that good text-to-speech no longer needs a large model. At 82 million parameters it is tiny, runs comfortably on a laptop CPU, and produces natural 24kHz speech that has scored surprisingly well in blind listening tests against much larger systems. It ships a set of fixed voices rather than cloning a target speaker, which suits narration, notifications, and accessibility, where a consistent, pleasant voice matters more than mimicry. Apache 2.0 licensed, it is an easy, dependency-light way to add speech output to a local pipeline. For voice cloning, look at a model like Chatterbox instead.

Strengths

  • Very small and fast, runs comfortably on CPU
  • Natural-sounding output for the size, with a range of built-in voices
  • Apache 2.0, so safe to use commercially without caveats

Weaknesses

  • Fixed built-in voices, not zero-shot voice cloning
  • Fewer languages than some larger TTS models
  • Less expressive control than heavier models built for emotion and style

Hardware requirements

QuantisationApprox. VRAMNotes
FP16~0.3GBFull precision, trivial to run on CPU or any GPU

Also runs on CPU (slower). Optimised builds available for Apple Silicon.

What you'd need to run this

Roughly what a machine to run this would need, at up to three levels of quality. Memory is the deciding factor.

Minimum to run it

FP16 · ~0.3GB needed

One 12GB GPU

NVIDIA GeForce RTX 3060 12GB

or a Mac or mini-PC with unified memory, if you prefer no discrete GPU, Mac mini M5 Pro .

At least 32GB of system RAM alongside the card.

around £700–£1,100

What else 12GB runs →

Licence

Apache 2.0 read the licence

Kokoro 82M: common questions

What hardware do I need to run Kokoro 82M?
At its most compressed (FP16) it needs roughly 0.3GB of VRAM, and about 0.3GB for good quality. VRAM figures are approximate and depend on context length and settings.
Is Kokoro 82M free for commercial use?
Yes. Kokoro 82M is licensed under Apache 2.0, which permits commercial use with no meaningful conditions.
Can I run Kokoro 82M on Apple Silicon?
Yes. Kokoro 82M has builds optimised for Apple Silicon, through MLX or GGUF on a Mac.
Does Kokoro 82M run on CPU?
Yes, Kokoro 82M can run on the CPU, though generation is slower than on a GPU.

Availability

Recommended for

  • Fast, clean narration on modest hardware
  • Adding local text-to-speech to an app without a big model
  • Commercial use where a permissive licence matters

Run it with

Related guides

Glossary

Catalogue entry last verified 30 July 2026. Specifications change; verify anything you are about to spend money on.