Updated 2026-10-15 · ranked by real download data
Best open-source TTS
Ranked nightly from our download snapshots of the Hugging Face catalog. Every entry shows its licence and what hardware it realistically needs.
| # | Model | Params | Context | Commercial use | 30d | Min VRAM |
|---|---|---|---|---|---|---|
| 01 | Kokoro-82M | 82 M | — | ✓ apache-2.0 | 10.5 M | from 0.7 GB |
| 02 | XTTS-v2 | — | — | other | 6.4 M | — |
| 03 | audio.cpp-gguf | — | — | other | 6.2 M | from 1.6 GB |
| 04 | Qwen3-TTS-12Hz-1.7B-CustomVoice | 1.9 B | — | ✓ apache-2.0 | 2.2 M | from 5.8 GB |
| 05 | chatterbox | — | — | ✓ mit | 1.6 M | from 12.2 GB |
| 06 | OmniVoice | 610 M | — | unknown | 1.5 M | from 4.2 GB |
| 07 | VieNeu-TTS-v3-Turbo | 131 M | 1 K | ✓ apache-2.0 | 1.1 M | from 0.7 GB |
| 08 | Qwen3-TTS-12Hz-0.6B-CustomVoice | 910 M | — | ✓ apache-2.0 | 1.0 M | from 3.4 GB |
| 09 | kokoro-inno-clone-tuner | 9.7 M | — | ✓ apache-2.0 | 977 K | from 0.5 GB |
| 10 | F5-TTS | — | — | ✗ cc-by-nc-4.0 | 820 K | from 5.0 GB |
| 11 | VoxCPM2 | 2.3 B | — | ✓ apache-2.0 | 710 K | from 5.9 GB |
| 12 | Kokoro-82M-v1.0-ONNX | 82 M | — | ✓ apache-2.0 | 690 K | from 0.7 GB |
| 13 | VibeVoice-1.5B | 2.7 B | — | ✓ mit | 604 K | from 6.9 GB |
| 14 | Qwen3-TTS-12Hz-0.6B-Base | 910 M | — | ✓ apache-2.0 | 570 K | from 3.4 GB |
| 15 | sanoTTS | — | — | gpl-3.0 | 504 K | from 0.5 GB |
| 16 | higgs-tts-2-3b-base | 5.8 B | 2 K | other | 415 K | from 14.1 GB |
| 17 | Qwen3-TTS-GGUF | — | — | ✓ apache-2.0 | 389 K | from 1.8 GB |
| 18 | Qwen3-TTS-12Hz-1.7B-VoiceDesign | 1.9 B | — | ✓ apache-2.0 | 362 K | from 5.8 GB |
| 19 | Higgs-Audio-v3-Studio | — | — | other | 342 K | from 5.0 GB |
Membership and ranking refresh nightly after the snapshot run. VRAM is an estimate for the smallest quantization at 8K context — methodology.