Africa TTS

Text-to-speech for 92 African languages and English, with voice cloning, built on Kyutai's Pocket TTS. Every language has a default voice, so all you need is text and a language code.

🎙️ Try it live · 🔊 Samples · 🤗 Model

An ongoing project: native-speaker feedback is what improves it most. See CONTRIBUTING.md.

⚡ Fast and high-fidelity speech on CPU, no GPU needed

Africa TTS runs on PyTorch (fp32 unquantised) by default: fast, full-fidelity audio without quantization degradation, faster than real time on an ordinary CPU. On a 4-vCPU cloud machine it speaks 2.4× faster than real time (speed benchmark). It also runs on an NVIDIA GPU with device="cuda".

🎯 Intelligible

On held-out sentences transcribed by Omnilingual ASR, Africa TTS is more intelligible than MMS-TTS in 21 of 27 languages. The 813M OmniVoice, seven times its size, is more intelligible in 24 of 29, by small margins: benchmark.

Install

pip install africa-tts

English also needs espeak-ng (apt install espeak-ng / brew install espeak-ng). The first time, it asks before downloading the model (about 440 MB) from Hugging Face, then caches it and works offline. In scripts, approve the download with --yes, AfricaTTS.load(download=True), or africa-tts --download beforehand. On a CPU-only Linux machine, you can install the CPU build of PyTorch with pip install --extra-index-url https://download.pytorch.org/whl/cpu africa-tts.

Use

africa-tts "Akwaaba! Yɛfrɛ me Kofi." --lang twi -o twi.wav
africa-tts --list-languages
africa-tts serve                       # the API server below, on port 8000

from africa_tts import AfricaTTS

tts = AfricaTTS.load()                                                  # PyTorch, CPU (default)
audio = tts.generate("Akwaaba! Yɛfrɛ me Kofi.", lang="twi")              # default Twi voice
audio = tts.generate("Hello Ghana!", lang="eng", voice="my_voice.wav")  # clone a 3-10 s clip
tts.save(audio, "out.wav")

# Code-switching: English loanwords & entities are spoken in English (enabled by default)
audio_cs = tts.generate("Mepɛ sɛ mekɔ high school wɔ regional hospital no nkyɛn.", lang="twi")

for chunk in tts.stream("Akwaaba! Yɛfrɛ me Kofi.", lang="twi"):         # 80 ms chunks
    ...

tts_gpu = AfricaTTS.load(device="cuda")                                 # PyTorch on NVIDIA GPU

• Languages: ISO 639-3 codes (twi, hau, yor, swh, zul, amh, eng, ...); see africa-tts --list-languages.
• Code-switching: In multilingual speech, African languages frequently borrow English nouns, technical terms, and named entities. Africa TTS automatically detects English words and phrases (curated and language-ID-verified from AfriSpeech/africa-nouns and AfriSpeech/african-named-entities) and pronounces them with English phonetics (espeak-ng) while preserving the surrounding words in native African phonology. Disable with codeswitch=False.
• Numbers: spell them out ("twenty", not "20").

Run it elsewhere

More

• Benchmarks: speed, and intelligibility per language
• Model, how it was made, data and limitations
• Training recipe · ONNX export · training code
• Status and roadmap · Contributing

License

Code: MIT (LICENSE). Model and default voices: CC-BY-NC-4.0. espeak-ng (English) is GPL-3.0-or-later. Built by AfriSpeech.