All models
voice-cloning

XTTS v2

CPML (non-commercial)

Voice cloning from a short reference clip across 17 languages

XTTS v2 is the flagship model of the Coqui TTS framework and one of the most widely used open voice-cloning models ever released. It clones a voice from a ~6-second sample and speaks 17 languages, with cross-language cloning — record in English, speak in Japanese. Coqui the company shut down, but the model remains a community staple.

Hear XTTS v2

XTTS v2 · Reference clone · Same three scripts across the directory.

Run it yourself

Best for: Cross-language cloning from a short reference recording.

Noncommercial / separate license: The CPML restricts both this model and its outputs to noncommercial use. Code licensing does not remove the weight restriction.

Code: MPL-2.0. Weights: Coqui Public Model License 1.0. License sources checked 2026-09-20. Source 1Source 2Source 3Source 4

Before you choose: The weights use CPML with non-commercial restrictions. Review the model license before choosing it for a product.

The catalog’s 6 GB figure is a rough historical estimate, not a measured minimum. CUDA is useful for practical inference; follow the maintained Coqui TTS instructions.

Details reviewed 2026-09-14. Hardware figures are estimates unless a benchmark is linked.

Compare with similar

Compare all →
Spark-TTSListen side-by-side →OpenVoice v2Listen side-by-side →F5-TTSListen side-by-side →

Choose an alternative to hear both models together.

Voices

1 voice
Female
Reference clonedefault
american

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.”

Run XTTS v2

pip install TTS

First audio example

Based on the author’s API. Install the documented dependencies and choose a compatible Python environment first.

# Use the Python environment described in the Coqui documentation.
# Provide a clean reference clip that you have permission to use.
from TTS.api import TTS
tts = TTS("tts_models/multilingual/multi-dataset/xtts_v2").to("cuda")
tts.tts_to_file(text="Hello from OpenSpeech.",
    speaker_wav="reference.wav", language="en", file_path="speech.wav")

Sources & setup details

XTTS v2 FAQ

Do I need a GPU to run XTTS v2?

The directory lists roughly 6 GB of VRAM as a historical estimate. Check the model documentation for your checkpoint and runtime.

Can XTTS v2 clone voices?

Yes — XTTS v2 supports voice cloning from reference audio.

What license is XTTS v2 released under?

XTTS v2 is released under the CPML (non-commercial) license. Always check the repository for the exact terms — some models license code and weights separately.

What languages does XTTS v2 support?

XTTS v2 supports English, Spanish, French, German, Italian, Portuguese, Polish, Turkish, Russian, Dutch, Czech, Arabic, Chinese, Japanese, Hungarian, Korean, Hindi.

Is there a hosted XTTS v2 API?

OpenSpeech Cloud is accepting production-access interest for XTTS v2, but this model is not currently hosted in Cloud. Use the official repository for current deployment options.