TTS voice files

Copies of third-party text-to-speech voice files, some converted for browser use. Each file is named by its SHA-256 hash.

File Family What it is Size Licence
320564d2615f….onnx Kitten Shared Kitten Nano model (unchanged) 56.77 MB Apache-2.0
6178872b9b7f….npz Kitten Bruno voice (one voice split from voices.npz) 0.41 MB Apache-2.0
af32a058886d….npz Kitten Luna voice (one voice split from voices.npz) 0.41 MB Apache-2.0
c8f914896f50….npz Kitten Jasper voice (one voice split from voices.npz) 0.41 MB Apache-2.0
a6a6855a5ecb….npz Kitten Rosie voice (one voice split from voices.npz) 0.41 MB Apache-2.0
c893d7b9e409….npz Kitten Hugo voice (one voice split from voices.npz) 0.41 MB Apache-2.0
5293bccd9cce….npz Kitten Kiki voice (one voice split from voices.npz) 0.41 MB Apache-2.0
c674edbf286e….npz Kitten Leo voice (one voice split from voices.npz) 0.41 MB Apache-2.0
884e56a48794….npz Kitten Bella voice (one voice split from voices.npz) 0.41 MB Apache-2.0
84602f9eceb4….onnx Pocket Shared Pocket model file (converted for the browser runtime) 111.86 MB CC-BY-4.0
520be4885a62….onnx Pocket Shared Pocket model file (converted for the browser runtime) 22.62 MB CC-BY-4.0
a717f6fc3002….onnx Pocket Shared Pocket model file (converted for the browser runtime) 16.39 MB CC-BY-4.0
43903126bc7f….npy Pocket Shared Pocket model file (converted for the browser runtime) 0.00 MB CC-BY-4.0
a9ec15e02043….json Pocket Shared Pocket model file (converted for the browser runtime) 0.12 MB CC-BY-4.0
5c8dca205b61….kv Pocket Jane voice state (converted from the Kyutai voice embedding) 7.37 MB CC-BY-4.0
a3f7679b6a6e….emb Pocket Jane voice embedding 0.00 MB CC-BY-4.0
6fcef9e14af2….kv Pocket Azelma voice state (converted from the Kyutai voice embedding) 7.96 MB CC-BY-4.0
a3f7679b6a6e….emb Pocket Azelma voice embedding 0.00 MB CC-BY-4.0
178f8ce7a382….kv Pocket George voice state (converted from the Kyutai voice embedding) 6.24 MB CC-BY-4.0
a3f7679b6a6e….emb Pocket George voice embedding 0.00 MB CC-BY-4.0
565111456dac….kv Pocket Michael voice state (converted from the Kyutai voice embedding) 7.28 MB CC-BY-4.0
a3f7679b6a6e….emb Pocket Michael voice embedding 0.00 MB CC-BY-4.0

Sources and changes

Kitten (Apache-2.0, KittenML)

  • From KittenML/kitten-tts-nano-0.8-fp32, revision 7a1db645b1f3ab9420761d87428e042b9cec3f26.
  • Model unchanged. voices.npz split into one file per voice; arrays unchanged.

Pocket (CC-BY-4.0, Kyutai)

  • Model from kyutai/pocket-tts, english_drifting_26-09, converted to browser ONNX graphs with INT8 weights by PocketTTS-RAVEN, built on PocketTTS.cpp.
  • Voices from kyutai/pocket-tts-without-voice-cloning, revision 1e08e6a23401048648a9fdcfde2f89348215c2a7, converted to saved model states (.kv).
  • Use must follow Kyutai's terms in the LICENSE-KYUTAI-* files.

Licences: LICENSES.md.

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support