kitten-tts-micro

A balanced KittenTTS model at 40M parameters. Good quality with a smaller footprint, suitable for most applications.

kitten-tts-micro sits between mini and nano in the KittenTTS model family, offering a good balance between audio quality and model size.

Specifications

Property	Value
Model ID	KittenML/kitten-tts-micro-0.8
Parameters	40M
Size	41 MB
Quality	Balanced

When to use this model

• Most general-purpose applications
• When you want good quality without the full footprint of kitten-tts-mini
• Server-side deployments with moderate storage or memory constraints
• Default choice when you are unsure which model to start with

kitten-tts-micro is a good starting point for most projects. You can always switch to kitten-tts-mini if quality is not sufficient, or kitten-tts-nano if you need faster inference.

Usage

from kittentts import KittenTTS
import soundfile as sf

model = KittenTTS("KittenML/kitten-tts-micro-0.8")
audio = model.generate(
    "Welcome to Kitten TTS, the ultra-lightweight text-to-speech library.",
    voice="Jasper"
)
sf.write("output.wav", audio, 24000)

The model is downloaded from Hugging Face on first use and cached locally. Pass cache_dir to control where it is stored.

Related models

• kitten-tts-mini — 80M params, highest quality
• kitten-tts-nano — 15M params, fastest inference