← All episodes
EP01: I Cloned My Voice With 8 AI Engines — Here's What Won
ElevenLabs, Cartesia, Coqui and 5 more voice cloning engines tested head to head. Audio demos, cost breakdown, and blind A/B test results.
Show Notes
What We Tested
Eight voice cloning engines evaluated across quality, cost, training data requirements, and multilingual support:
- ElevenLabs
- Cartesia
- Coqui / XTTS
- LMNT
- Fish Audio
- StyleTTS2
- OpenAI
- Deepgram
Key Findings
- The open-source option (Coqui XTTS) needed just 5 seconds of audio
- The winner (Cartesia) needed 54 minutes but produced a clone that fooled colleagues in blind tests
- Cost ranged from free (open source) to $99/month (enterprise)
- Multilingual support varied wildly — only 2 engines handled German well
Timestamps
- 00:00 — Introduction & credentials
- 02:07 — AI voice clone bridge (Cartesia demo)
- 02:59 — NotebookLM deep dive begins
- 42:00 — Key takeaways
- 45:26 — Outro & next episode preview
Links
Transcript coming soon. Until then, the companion article covers everything we talked about in this episode.
Subscribe to the podcast
Pick your platform and you'll get every new episode automatically.