A fully-offline, local studio for multiple open-source TTS engines — VibeVoice, Kokoro, Chatterbox, OmniVoice, VoxCPM & Qwen3-TTS. It also supports Voice-to-Text with Whisper
Framer AI AgentsDesign and publish professional sites with AI
Promoted
Maker
📌
Hey Product Hunt! 👋 Maker here.
As a content creator for 15 years (YouTube/@WPAcademyPK), I often need voiceovers for my video tutorials, courses, and product demos. So, I used to go to ElevenLabs or Magnific to generate voices that are truly human-like and fit my use-case. But they are expensive, especially when the audios are hours long.
So I built Voice Studio, a local, fully-offline studio for open-source voice models that runs entirely on your own machine. Initially, I built this for my personal use, but later on I thought this could help so many creators like me, so here it is. It is now available on GitHub, and I will push more updates in the future.
What you can do:
🎙️ Text-to-Voice across 6 open-source engines (VibeVoice, Kokoro, Chatterbox, OmniVoice, VoxCPM, Qwen3-TTS). Swap between them in one click.
🧬 Clone a voice from a single short sample
🎧 Build multi-speaker podcasts line by line
📝 Transcribe any audio with Whisper: 99 languages, editable text, exportable SRT/VTT subtitles
The part I care most about: once the models download, it works with no internet, no API keys, and no per-word billing. Your voice never leaves your machine. It's free and open-source. An NVIDIA GPU helps, but it runs on CPU too.
It's early, and I'm building in the open. I'd genuinely love your feedback on which engines feel best, what's confusing, and what you'd want next. Every comment gets a reply. 🙏
Mian Shahzad Raza (MSR)
Report
would love a built-in voice cloning feature so i can record a short sample and have the studio generate speech in my own voice across all the supported engines
Report
Maker
@hatunkeski92266 That's fully implemented, you can upload yor own custom voices in the left panel.
Report
Would love to see a built-in voice preset sharing option, like being able to export a fine-tuned voice config as a small file so other users can drop it in and get the same result. Would make collaboration way easier.
Voice cloning preview would be huge — let users hear a side-by-side comparison between the original voice and the cloned version before exporting. Currently it's hard to know if the clone captured the right tone without running the full generation first.
Report
finally a tts studio that actually works offline without weird setup hoops, kokoro sounds surprisingly natural and the whisper transcription is honestly better than i expected
the fact that everything runs fully offline with so many engines bundled into one clean studio is honestly impressive, love that whisper transcription is included too
Report
Maker
@ekindzalannnbs Tried my best, spent weeks fine tuning it. More engines, models, and features will come soon.
would love a built-in voice cloning feature so i can record a short sample and have the studio generate speech in my own voice across all the supported engines
@hatunkeski92266 That's fully implemented, you can upload yor own custom voices in the left panel.
Would love to see a built-in voice preset sharing option, like being able to export a fine-tuned voice config as a small file so other users can drop it in and get the same result. Would make collaboration way easier.
@rbecerir62672 That's a geat idea, can you open an issue on the GitHub? https://github.com/msrbuilds/voice-studio/issues
Voice cloning preview would be huge — let users hear a side-by-side comparison between the original voice and the cloned version before exporting. Currently it's hard to know if the clone captured the right tone without running the full generation first.
finally a tts studio that actually works offline without weird setup hoops, kokoro sounds surprisingly natural and the whisper transcription is honestly better than i expected
@fcalapverd79575 Gald to know that it works for you :)
Pazi
Congrats on the Voice Studio launch! It looks super promising and exciting. 🚀
@zvonimir_sabljic1 Thank you 🙌
the fact that everything runs fully offline with so many engines bundled into one clean studio is honestly impressive, love that whisper transcription is included too
@ekindzalannnbs Tried my best, spent weeks fine tuning it. More engines, models, and features will come soon.