neuvoice 1.0.0

Fast, on-device neural TTS by Lamapi

---

license: openrail language:

• tr
• bg
• el
• fr
• ko
• ja
• ar
• de
• da
• cs
• es
• et
• fi
• hr
• hu
• id
• it
• lt
• lv
• nl
• pl
• pt
• ro
• ru
• sk
• sl
• sv
• uk
• vi
• na tags:
• text
• speech
• tts
• neuvoice
• supertonic

---

Neuvoice — Fast, On-Device Neural TTS by Lamapi

Neuvoice is a lightweight, on-device text-to-speech library that runs entirely via ONNX Runtime — no cloud calls, no latency surprises. Built by Lamapi, it ships with a curated voice library and supports 31 languages out of the box.

Quick Start

pip install neuvoice

from neuvoice import TTS

tts = TTS(auto_download=True)
style = tts.get_voice_style("Alina")

text = "Hello! Welcome to Neuvoice — fast, private, on-device speech synthesis."
wav, duration = tts.synthesize(text, voice_style=style, lang="en")

tts.save_audio(wav, "output.wav")
print(f"Generated {duration:.2f}s of audio")

On first run, model assets are downloaded and cached automatically under ~/.cache/neuvoice.

Highlights

31 supported languages. Covers a wide range of scripts and regions, from European languages to Arabic, Hindi, Japanese, Korean, Vietnamese, and more.

Runs entirely on-device. ONNX Runtime powers inference — no API keys, no network dependency after the initial model download. CPU is sufficient; GPU acceleration is supported when available.

Rich voice library. Ships with voices including Alina, Cem, Cole, Giray, Leon, Lina, Linda, Mustafa, Sarp, Selin, Sema, and Soras. Each voice is a compact style embedding — load any of them by name in a single call.

Inline expression tags. Embed <happy>, <laugh>, <breath>, <sad>, and other tags directly in your text to shape the delivery without any extra parameters.

Long-form synthesis. Inputs are automatically chunked, synthesized, and rejoined with configurable silence — no manual splitting required.

Supported Languages

Code	Language	Code	Language	Code	Language	Code	Language
en	English	ko	Korean	ja	Japanese	ar	Arabic
bg	Bulgarian	cs	Czech	da	Danish	de	German
el	Greek	es	Spanish	et	Estonian	fi	Finnish
fr	French	hi	Hindi	hr	Croatian	hu	Hungarian
id	Indonesian	it	Italian	lt	Lithuanian	lv	Latvian
nl	Dutch	pl	Polish	pt	Portuguese	ro	Romanian
ru	Russian	sk	Slovak	sl	Slovenian	sv	Swedish
tr	Turkish	uk	Ukrainian	vi	Vietnamese	na	(fallback)

Available Voices

tts = TTS()
print(tts.voice_style_names)
# ['Alina', 'Cem', 'Cole', 'Giray', 'Leon', 'Lina', 'Linda',
#  'Mustafa', 'Sarp', 'Selin', 'Sema', 'Soras']

Load any voice by name:

style = tts.get_voice_style("Selin")

Or from a custom style file:

style = tts.get_voice_style_from_path("/path/to/my_voice.json")

synthesize() Parameters

Parameter	Type	Default	Description
text	str	—	Input text. Supports inline tags like <happy>.
voice_style	VoiceStyle	—	Voice loaded via get_voice_style().
total_steps	int	5	Flow-matching denoising steps. Higher = better quality, slower inference. Range: 1–100.
speed	float	1.05	Playback speed multiplier. Range: 0.7–2.0.
lang	str	"en"	ISO 639-1 language code, or "na" for unknown languages.
max_chunk_length	int	300	Max characters per synthesis chunk (120 for Korean).
silence_duration	float	0.3	Seconds of silence inserted between chunks.
verbose	bool	False	Print per-chunk progress to stdout.

Returns: (waveform, duration) — waveform as a (1, samples) NumPy array, duration in seconds.

Examples

Multilingual synthesis:

tts = TTS()
style = tts.get_voice_style("Leon")

pairs = [
    ("Merhaba! Bugün hava çok güzel.", "tr"),
    ("Bonjour! Il fait beau aujourd'hui.", "fr"),
    ("こんにちは！今日はいい天気ですね。", "ja"),
]
for text, lang in pairs:
    wav, dur = tts.synthesize(text, voice_style=style, lang=lang)
    tts.save_audio(wav, f"output_{lang}.wav")

Expression tags:

style = tts.get_voice_style("Lina")
text = "Good news! <happy> We just shipped the feature. <laugh> Don't tell anyone yet."
wav, dur = tts.synthesize(text, voice_style=style, lang="en")

Higher quality with more steps:

wav, dur = tts.synthesize(
    text="A slow, deliberate reading for maximum clarity.",
    voice_style=style,
    total_steps=30,
    speed=0.9,
    lang="en",
)

Configuration

Model cache location and thread counts can be controlled via environment variables:

Variable	Description
NEUVOICE_CACHE_DIR	Override the default cache directory (~/.cache/neuvoice).
NEUVOICE_MODEL_REPO	Override the Hugging Face model repository.
NEUVOICE_REVISION	Model revision/branch to use (default: main).
NEUVOICE_INTRA_THREADS	ONNX intra-op thread count (default: auto).
NEUVOICE_INTER_THREADS	ONNX inter-op thread count (default: auto).

License

Neuvoice is released under the MIT License. See LICENSE for details.

The bundled ONNX model is released under the OpenRAIL-M License.

Copyright © 2026 Lamapi