Bark
Suno's expressive transformer-based TTS.
Open Source 8–12 GB VRAM
- Min VRAM
- 8 GB
- GPU class
- Entry GPU
- Quant
- FP16
Actually FreeNo SignupOpen SourceWatermark-Free
Tortoise TTS
Slow, but the quality is worth the wait.
Open Source 4–8 GB VRAMRuns locally
Actually FreeNo SignupOpen SourceWatermark-FreeHobbyist-Friendly
Visit Tortoise TTSUpdated 2024-11-02 · Direct link
Hardware requirements
Runs locally · Entry GPU (6–8 GB)
Min VRAM
4 GB
Rec. VRAM
8 GB
Min RAM
8 GB
Rec. RAM
16 GB
Disk
5 GB
GPU class
Entry GPU
11.6+No Apple SiliconCPU-CapableQuant: FP16
Workable at 4 GB VRAM; 8 GB recommended. CPU is impractical.
Screenshot placeholder · Tortoise TTS
What is Tortoise TTS?
Tortoise is the high-quality, low-speed TTS that set the bar before XTTS landed. Multi-step diffusion-style generation produces remarkably natural prosody from just a few seconds of reference audio. Now mostly displaced by faster models, but still notable for clone fidelity on a budget.
Pros & cons
Pros
- Excellent prosody and naturalness vs. its 2022 contemporaries
- Voice cloning from a few seconds of reference
- Mature codebase with many community forks
Cons
- Glacially slow — minutes per sentence on consumer GPUs
- Newer models (XTTS, F5-TTS) match quality with 10–50× speedup
What's actually free?
Apache 2.0; weights and code both free.
✓ Actually FreeNo SignupOpen SourceWatermark-Free
Alternatives
XTTS v2 (Coqui)
Multilingual voice cloning in 6 seconds.
Open Source 4–6 GB VRAM
- Min VRAM
- 4 GB
- GPU class
- Entry GPU
- Quant
- FP16
Actually FreeNo SignupOpen SourceWatermark-Free
F5-TTS
Zero-shot voice cloning TTS — 15 s of audio is enough.
Open Source 8–12 GB VRAM
- Min VRAM
- 8 GB
- GPU class
- Mid GPU
- Quant
- FP16
Actually FreeNo SignupOpen SourceWatermark-Free