Hi Product Hunt 👋

I built MusaTTS because I wanted good AI voices without uploading every script to someone else's server and paying per character.

MusaTTS is a desktop app for Windows and Apple Silicon Macs. The voice engines run on your own machine, and an optional NVIDIA GPU pack on Windows makes them much faster.

What's in 5.0:
🎙️ Voice Studio: 46 built-in voices in 8 languages, with pauses, "Say as" for dates and phone numbers, and long scripts split into chunks automatically
🧬 Voice Clone Pro: clone a voice from a 3–10 second sample with four engines. Musa Solo runs on any computer, Musa Aria does expressive English with laughs and sighs, Musa Chorus covers 23 languages, and Musa Sona 600 covers 646 (needs an 8 GB GPU)
🎬 Dubbing: transcribe a video, translate it, and re-voice it in the original speaker's voice, with SRT import and export
🔁 Audio Clone: change the voice in an existing recording while keeping the words and timing
👥 Podcast: multi-speaker conversations from one script
🔌 Local API: an OpenAI-compatible speech endpoint, so your own tools can use it

A few honest notes:
- Paid plans need an online license check every 7 days. Generation itself happens on your computer.
- Dubbing and auto-transcribe use your own Groq or OpenAI API key.
- Requirements: Windows 10/11 64-bit with an AVX2 CPU and 8 GB RAM, or an M1-or-newer Mac on macOS 13+. Intel Macs aren't supported.

The free version is the trial: 10,000 characters a day in Voice Studio, 3,500 in Voice Clone Pro and 1 Audio Clone a day. Please try it on your own computer before buying.

I'd love your feedback, especially on which languages sound best or worst for you. I'll be here answering questions all day 🙏

Hi Product Hunt 👋

I built MusaTTS because I wanted good AI voices without uploading every script to someone else's server and paying per character.

MusaTTS is a desktop app for Windows and Apple Silicon Macs. The voice engines run on your own machine, and an optional NVIDIA GPU pack on Windows makes them much faster.

What's in 5.0:
🎙️ Voice Studio: 46 built-in voices in 8 languages, with pauses, "Say as" for dates and phone numbers, and long scripts split into chunks automatically
🧬 Voice Clone Pro: clone a voice from a 3–10 second sample with four engines. Musa Solo runs on any computer, Musa Aria does expressive English with laughs and sighs, Musa Chorus covers 23 languages, and Musa Sona 600 covers 646 (needs an 8 GB GPU)
🎬 Dubbing: transcribe a video, translate it, and re-voice it in the original speaker's voice, with SRT import and export
🔁 Audio Clone: change the voice in an existing recording while keeping the words and timing
👥 Podcast: multi-speaker conversations from one script
🔌 Local API: an OpenAI-compatible speech endpoint, so your own tools can use it

A few honest notes:
- Paid plans need an online license check every 7 days. Generation itself happens on your computer.
- Dubbing and auto-transcribe use your own Groq or OpenAI API key.
- Requirements: Windows 10/11 64-bit with an AVX2 CPU and 8 GB RAM, or an M1-or-newer Mac on macOS 13+. Intel Macs aren't supported.

The free version is the trial: 10,000 characters a day in Voice Studio, 3,500 in Voice Clone Pro and 1 Audio Clone a day. Please try it on your own computer before buying.

I'd love your feedback, especially on which languages sound best or worst for you. I'll be here answering questions all day 🙏

Musa TTS

Offline AI Text to Speech & Voice Cloning Software

Offline AI Text to Speech & Voice Cloning Software