F5-TTS
Official code for "F5-TTS: A Fairytaler that Fakes Fluent and Faithful Speech with Flow Matching"
F5-TTS scores 6.3/10 (Solid) in Voice & Realtime.
Language Python License MIT Stars 14.9k★ Type model Self-host Yes Repo SWivid/F5-TTS Homepage https://swivid.github.io/F5-TTS/
Quality
Solid 6.3
Adoption 54
Activity 42
Maturity 86
Community 59
Capability 98
Show the math
Overall 6.3/10
value = 6.3/10
score = value
Adoption 5.9/10
value = 54.3/100
score = 1 + 9 * value / 100
Activity 4.7/10
value = 41.5/100
score = 1 + 9 * value / 100
Maturity 8.8/10
value = 86.3/100
score = 1 + 9 * value / 100
Community 6.3/10
value = 59/100
score = 1 + 9 * value / 100
Capability 9.8/10
value = 97.6/100
score = 1 + 9 * value / 100
Key metrics
14.9k★ GitHub stars
9/90d Recent commits
4 devs/90d Recent contributors
2.42% WER WER, LibriSpeech-PC test-clean (zero-shot TTS intelligibility) F5-TTS paper reports 2.42% WER on its released LibriSpeech-PC test-clean subset (32 NFE + Sway Sampling), measuring intelligibility of generated speech; reported as 100-WER. Source: arXiv 2410.06885 (https://arxiv.org/html/2410.06885v1).
1.7y Project age
2mo ago Last commit
Gotchas
No gotchas documented yet. Contribute one if you know a constraint we missed.
Labels
- Self-hosted
Self-hosted
You deploy and operate it yourself; there is no hosted option here.
tts flow-matching zero-shot voice-cloning diffusion