Which AI Models Does TTS WebUI Support?

Updated

TTS WebUI supports more than 30 AI models for text-to-speech, audio, and music generation. The site names Bark, MusicGen, Tortoise, and Vocos as examples, and describes the platform as a free, Gradio-based web interface. If you want one interface that can route different audio tasks to different models rather than installing a separate tool for each, that breadth is the main reason to use it. The exact full model list is not published on the page, so treat "30+" as the supported scale and the four named models as confirmed examples.

Confirmed models and what they are suited to

Model Task area Why you would pick it
Bark Text-to-speech Speech generation with access to a wide range of voices, including additional voices from the Bark Speaker Directory
MusicGen Music generation Producing music audio rather than spoken narration
Tortoise Text-to-speech An alternative TTS model when you want a different voice or delivery than Bark produces
Vocos Audio generation / vocoding A supporting model in the audio pipeline rather than a standalone TTS front end

The page groups these under "versatile AI models" used for "different TTS tasks," so the practical split is: Bark and Tortoise for speech, MusicGen for music, Vocos for audio processing. Because the site does not publish a per-model capability matrix, verify a specific model's behavior inside the interface before committing a long project to it.

How to match a model to your task

  1. Decide the output type first. Spoken narration points to Bark or Tortoise; music points to MusicGen.
  2. Pick a voice source for speech. TTS WebUI offers a variety of voices plus additional voices from the Bark Speaker Directory, so if voice variety is your priority, start with Bark.
  3. Generate a short sample. The site describes generation as "just a few clicks," so test a sentence or two before producing a full track or long narration.
  4. Compare models on the same text. If Bark's output does not fit, run the identical input through Tortoise to hear the difference.
  5. Confirm the result. You should get playable audio from the selected model; if not, the model or its dependencies are the first thing to check.

Installation choices that affect which models you can run

TTS WebUI offers two installation paths, and your choice can matter for model availability and performance:

  • One-click installers — the faster route if you want to start generating audio without managing dependencies.
  • Manual installation — more control, and the better option if you need to adjust the environment for specific models.

The page also points to documentation alongside the installation options, so check that before assuming a given model is included in your setup.

What the page does not tell you

  • The complete list of all 30+ models is not published, only the four named examples.
  • No pricing details, account requirements, or usage limits appear on the page. The site calls the interface free, but that statement covers the software itself, not any third-party model or service you might connect.
  • No per-model hardware requirements, languages, or output formats are listed.

For anything beyond the four confirmed models, the GitHub repository and the documentation are the places to look, since the site directs users there for source code and setup details.

Bottom line

Choose TTS WebUI if you want a single Gradio interface covering speech, audio, and music with a choice of models. Start with Bark for voice variety, MusicGen for music, and Tortoise as your second speech option, then confirm the full model list in the documentation before building a workflow around a specific one.