Leokani-tts 7M
Leokani-tts is a small, fast English text-to-speech model: frontier-quality speech on any CPU, no GPU needed. It speaks in one female American English voice and was trained on synthetic speech from Kokoro-82M. The name is Hawaiian: leo (voice) and kani (sound).
Leokani-tts is from Milliwatt Labs: AI measured in milliwatts, not megawatts. We build small, efficient models that run on the devices you already own, with no data center and no GPU.
Apps for Mac and Android, a command-line tool for Linux and Raspberry Pi, and a JavaScript library are on milliwattlabs.com.
Samples
| Text | Audio |
|---|---|
| Hello! I'm a small text-to-speech model, and I run on an ordinary computer. | |
| That sounds like a great plan! Let me know how it goes, and good luck tomorrow. | |
| Did you remember to water the tomatoes before the rain started? | |
| Your flight leaves on March 14th at 7:45 AM, and the fare is 219 dollars. | |
| The trail is 4.2 miles long, climbs about 1,200 feet, and takes around 3 hours. | |
| Maria Gonzalez and Hiroshi Tanaka are meeting Dr. Okafor in Minneapolis on Friday. | |
| The old lighthouse stood at the edge of the bay for more than a century, and every evening the keeper climbed one hundred and twelve steps to light the lamp. | |
| Sure, no problem. |
Usage
hf download milliwattlabs/leokani-tts-7M --local-dir leokani-tts
cd leokani-tts
pip install -r requirements.txt # torch, safetensors
python leokani.py "Hello there. How was your day?" -o hello.wav
python leokani.py "A little slower, please." --speed 0.85 -o slow.wav
python leokani.py --fp16 "Same voice, half-size weights." -o fp16.wav
Leokani-tts needs espeak-ng to turn text into phonemes. espeak-ng is a separate dependency under GPL-3.0, and it
is not included here. Install it with brew install espeak-ng (macOS) or sudo apt install espeak-ng
(Debian/Ubuntu), or point to it with --espeak PATH. To skip it, pass Kokoro-style phonemes with
--phonemes "...".
Usage (ONNX)
pip install onnxruntime
python leokani_onnx.py "Hello there. How was your day?" -o hello.wav
python leokani_onnx.py --int8 "Same voice, the smallest file." -o int8.wav
INT8: ~half the size of FP16; rated the same as full precision in blind listening tests.
Specs
7.0M parameters 路 28 MB (fp32) or 14 MB (fp16) safetensors 路 leokani-tts.onnx (fp32), leokani-tts-fp16.onnx
(fp16) and leokani-tts-int8.onnx (int8, 7.8 MB) 路 24 kHz mono output.
Performance
| Model | Params | Size | Speed, Mac | Speed, 1 Arm core | First speech, Mac | First speech, 1 Arm core | WER | UTMOS |
|---|---|---|---|---|---|---|---|---|
| Leokani-tts 鹿 | 7.0M | 28 MB | 90脳 | 25脳 | 50 ms | 190 ms | 0.07 % | 3.53 |
| KittenTTS nano 0.8 | 14.0M | 60 MB | 26脳 | 3.6脳 | 240 ms | 1.7 s | 0.41 % | 2.18 |
| Piper (en_US lessac medium) | 15.7M | 63 MB | 49脳 | 5.6脳 | 86 ms | 710 ms | 0.90 % | 3.77 |
| KittenTTS mini 0.8 | 73M | 82 MB | 16脳 | 3.4脳 | 330 ms | 1.6 s | 0.21 % | 2.04 |
| Kokoro-82M | 82M | 328 MB | 7.6脳 | 0.48脳 | 670 ms | 11 s | 0.35 % | 3.76 |
| Kyutai Pocket TTS | 110M | 226 MB | 9.7脳 | 1.2脳 | 34 ms | 280 ms | 0.35 % | 3.25 |
| Chatterbox | 798M | 3.2 GB | 0.28脳 | 0.028脳 | 15 s | 150 s | 0.55 % | 3.36 |
鹿 PyTorch. The ONNX version gives the same output, within 1 % of this speed on the Arm core and 4 % on the Mac.
On an iPhone 17 Pro Max (Core ML, Neural Engine), Leokani-tts renders the 8 sample lines at 94脳 real time, 74 ms per line (median, the whole line). Kokoro-82M on the same phone takes 308 ms to its first audio chunk (median).
Speed = seconds of speech per second of compute; first speech = median time to first audio over 20 sentences. Mac: M3 Pro CPU; Arm: one 2.8 GHz Cortex-A725 core (Raspberry Pi-class). WER (Whisper large-v3), UTMOSv2: 120 sentences.
Limitations
- One voice, English only.
- Pronunciation follows espeak-ng, so unusual names may be misread. Some symbols are read literally ("$219" becomes "dollar two hundred nineteen"), so write amounts out ("219 dollars").
- Delivery is neutral, intended for a home-assistant style of inflection.
License
CC BY-NC 4.0: non-commercial use only (see LICENSE).
For a commercial license or a custom voice, email hello@milliwattlabs.com.
Training audio data generated from Kokoro-82M by hexgrad, which is licensed under
Apache-2.0 (licenses/Kokoro-82M-Apache-2.0.txt).
- Downloads last month
- -