VoiceStudio — open-source, fully-local ElevenLabs alternative
Self-contained local voice AI: voice cloning, voice design, video dubbing, dictation, transcription and audiobooks across 646 languages, 14 pluggable TTS engines.
Why I saved this#
Came from a Spanish-language X post hyping it as a free local ElevenLabs replacement. The framing is classic engagement bait (“ESTO ES JODIDAMENTE INCREÍBLE”), but the underlying repo checks out and is genuinely self-contained: voice stack that runs entirely on your own machine with no per-character billing, no usage caps, and no audio leaving the box.
Notes#
- Repo: debpalash/VoiceStudio , Python, AGPL-3.0, site https://voicestudio.sh
- Capabilities: voice cloning from a clean reference clip, voice design, video dubbing, dictation, transcription, audiobook generation; 646 languages claimed; pluggable TTS engines (14 per the tweet) instead of a single vendor engine.
- Verified at save time (2026-09-26 via GitHub API): 35.8K stars, 4.2K forks, not archived, pushed the same day — very active.
- Source tweet: https://x.com/liambraus/status/2103825720807993573 (@liambraus, ex-ElevenLabs CMO). Claims 19.4K stars — stale; the repo has since nearly doubled.
- AGPL-3.0: fine for personal/local use; viral license if it ever gets networked into a service.