Catalog
Datasets
The recordings behind the voices. See which datasets already have a model and which are ready to train.
130datasets
32With model
98To train
Browse
Full catalog575 voice models · 75 languagesCompare voicesThe same sentence, up to four voicesMy libraryYour favorite voices, on this deviceWhat's newAdded or updated in the last 14 daysBy language
All languages (75)Characteristics
Create a voice
Create an LJSpeech dataset for Piper TTSLJSpeech is the standard format expected by Piper. Each sentence is a separate WAV file, paired with its transcript in metadata.csv.Train a voice on KaggleKaggle assigns GPU quota to your account, and the allowance can vary with availability. Check the counter before starting a long session.Create a voice
Create an LJSpeech dataset for Piper TTSLJSpeech is the standard format expected by Piper. Each sentence is a separate WAV file, paired with its transcript in metadata.csv.Train a voice on KaggleKaggle assigns GPU quota to your account, and the allowance can vary with availability. Check the counter before starting a long session.Use a voice
Use a Piper voiceInstall Piper, get the two files for a voice, and generate a WAV locally. The commands below use the version maintained by the Open Home Foundation.Use Piper in a Node.js project with piperttspipertts brings Piper to Node.js, with file output or in-memory audio. The recommended mode is PiperNativeTTS: fully native in-process inference (ONNX plus a compiled espeak bridge), no Python, roughly 10x faster than the wrapper. Voices (.onnx plus .onnx.json) remain separate files.Piper Plus voicesSome voices use the Piper Plus architecture: multilingual synthesis, prosody, and MB-iSTFT decoding. They carry the Piper+ badge in the catalogue and require the piper-plus runtime, not the standard piper binary.Catalog
The recordings behind the voices. See which datasets already have a model and which are ready to train.