# KittenML/KittenTTS

State-of-the-art TTS model under 25MB 😻

Repository: https://github.com/KittenML/KittenTTS
Canonical: https://ross.abutalabs.com/products/kittentts
Language: Python
License: Apache-2.0
License Family: permissive
Last push: 2026-08-19T00:12:38+00:00

## Health v2 (maintenance only)
Score: 66/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 98, release rhythm 48, longevity 28
- inputs: {"age_days": 393, "days_push": 15, "days_rel": 190, "gap_med": 101.5, "n_releases_24m": 3}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 15403, forks 885 (observed 2026-08-28T04:11:09.792361+00:00)

## What it is
KittenTTS is an open-source, ultra-lightweight text-to-speech library built on ONNX, with models ranging from 15M to 80M parameters (25-80 MB on disk). It delivers high-quality 24 kHz voice synthesis on CPU without requiring a GPU, making it suitable for edge deployment.

## Use cases
- generate speech from text on a CPU without a GPU
- add text-to-speech to an edge device or embedded system
- run a tiny TTS model under 25MB
- synthesize voice audio for an app offline
- convert written text into natural-sounding speech in Python
- deploy lightweight voice synthesis on resource-constrained hardware

## When to choose
- you need TTS inference on CPU-only machines or edge devices
- model size and memory footprint are critical constraints
- you want a simple Python API with built-in voices and speed control
- you prefer an Apache-2.0 licensed open-source TTS solution

## When to avoid
- you need production-stable APIs - the project is a developer preview and APIs may change
- you need extensive voice cloning or many built-in voices (only 8 voices available)
- you require very high-fidelity studio-grade speech from large models
- you need non-ONNX runtimes or GPU-optimized training pipelines

## Facets
- artifact type: library
- maturity: active
- function: tts, speech-recognition, machine-learning, llm-inference
- domain: speech-processing, machine-learning, artificial-intelligence
- platform: python, cross-platform
- tags: onnx, lightweight, edge-deployment, voice-synthesis, cpu-inference, audio, cpu

## Member repositories
- KittenML/KittenTTS (main) score 66

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:11:09.792361+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:05:54.685988+00:00, confidence not recorded.
  - readme: https://github.com/KittenML/KittenTTS (fetched 2026-08-28T04:11:09.792361+00:00, sha 48c03c2b1202)
- Data as of 2026-08-30T08:39:29.467469+00:00.