All models
voice-cloning

Spark-TTS

Apache-2.0

Built on Qwen2.5, zero-shot voice cloning + style control

Spark-TTS builds speech generation directly on a 0.5B Qwen backbone with a single-stream codec, making it unusually simple and efficient. It does zero-shot cloning plus controllable voice creation — set gender, pitch, and speaking style to synthesize a new voice from scratch.

Voices

1 voice
Neutral
Defaultdefault
—
NeutralEmotionalNumbers & Dates

The scripts

Neutral
“The quick brown fox jumps over the lazy dog near the riverbank.”
Emotional
“I can't believe you actually did it. This is incredible!”
Numbers & Dates
“On March 14th, 2025, the team raised $4.2 million at a 38% margin.”

Install

pip install spark-tts

Spark-TTS FAQ

Do I need a GPU to run Spark-TTS?

For practical speeds, yes — plan on roughly 8 GB of VRAM. If you'd rather not manage a GPU, hosted access to every model in this directory is coming via OpenSpeech Cloud.

Can Spark-TTS clone voices?

Yes — Spark-TTS supports voice cloning from reference audio.

What license is Spark-TTS released under?

Spark-TTS is released under the Apache-2.0 license. Always check the repository for the exact terms — some models license code and weights separately.

What languages does Spark-TTS support?

Spark-TTS supports Chinese, English.

Is there a hosted Spark-TTS API?

Yes — OpenSpeech Cloud is launching hosted, pay-per-minute access to Spark-TTS and every other model in this directory. No CUDA, Docker, or GPU needed.