68
Stack Score
Editorial pick✓ Verified working
Coqui TTS
Open-source neural TTS toolkit (XTTS).
Try Coqui TTS✓ Free trial available
Best for
Engineers who want to self-host high-quality voice cloning without paying ElevenLabs per character.
At a glance
- Pricing
- free
- Setup time
- 30 minutes
- Learning curve
- advanced
- Last verified
- 5/22/2026
What we love
- ✓Best open-source voice quality available
- ✓No per-character cost
- ✓Strong multilingual
Where it falls short
- −No commercial support
- −Self-host requires GPU
- −CPML license restricts commercial use of base model
Pricing
✓ Pricing verified Jun 21, 20261 tier
Open Source
$0
forever
- ✓XTTS weights
- ✓All languages
- ✓Self-host
- ✓Community
Why we picked it
Coqui shut down the commercial product but XTTS lives on as the best open-weights voice cloning model. Engineering effort to deploy, but free.
Overview
Open-source neural TTS toolkit (XTTS). Engineers who want to self-host high-quality voice cloning without paying ElevenLabs per character.
Key features
- ●XTTS v2 open weights
- ●13 languages
- ●6-second voice cloning
- ●Streaming inference
- ●Apache + CPML license
- ●Local GPU inference
- ●Active community forks
Best use cases
Self-hosted dubbing
Clone a voice once, dub long-form content offline.
On-device TTS
Embed XTTS in a desktop app for offline use.
Integrations
Similar tools
See all voice & audio tools →One AI Stack pick every week.
The tool we’d pay for this week, why we’d pay for it, and what we’d skip. No promo spam.