Skip to content

Usage Overview

This section explains the everyday ways to use OpenTalking: start and debug services from the command line, then use WebUI to select avatars, choose models, configure voices, and validate a session.

If you have not started the project yet, begin with Quick Start. If WebUI already opens, this section is the next place to look.

Who This Section Is For

This section is for developers and integrators who have completed the basic setup and want to use OpenTalking more effectively. It covers:

  • Starting OpenTalking services, the frontend, and helper scripts.
  • Using WebUI for avatar, model, voice, and session configuration.
  • Preparing custom avatars and previewing or cloning voices.
  • Common parameters, ports, backends, and environment files.

It keeps model support, scenario tutorials, and API schemas out of the main usage flow so the page stays focused on how to use the current version.

Main Usage Modes

Command Line Tools

Command line tools are best for development, integration, deployment validation, and automation. Use them to start unified mode, select inference backends, set ports, prepare avatar assets, and run QuickTalk benchmarks.

Start with Command Line Tools, then use Advanced CLI Arguments when you need backend, port, host, or remote inference options.

WebUI

WebUI is best for interactive validation. It provides avatar selection, model selection, TTS provider and voice configuration, text or voice conversation, and status feedback.

Start with WebUI Basic Usage, then continue to Custom Avatar or Voice and TTS when needed.

I Want to Try the UI

  1. Start Mock mode from Quick Start.
  2. Open WebUI.
  3. Follow WebUI Basic Usage to select avatar, model, voice, and create a session.

Mock mode does not require model weights, so it is the fastest way to verify the UI and backend flow.

I Want to Customize Avatar

  1. Validate one built-in avatar first.
  2. Read Custom Avatar to understand image, video, and model requirements.
  3. Upload an image in WebUI, or prepare Wav2Lip assets with scripts.
  4. Select the new avatar in WebUI and test it with a short message.

I Want to Configure Voice / TTS

  1. Read Voice and TTS and choose the TTS provider.
  2. Select or preview a default voice in WebUI.
  3. If voice cloning is needed, prepare provider credentials, sample audio, and public access settings.

I Want to Start Services from CLI

  1. Read Command Line Tools for the main commands and scripts.
  2. Use scripts/start_unified.sh for Mock, local model, or OmniRT modes.
  3. Use Advanced CLI Arguments when you need custom ports, host binding, or environment files.

What This Section Does Not Cover

Business Examples

Customer-service avatars, live commerce, course narration, and private assistants belong in Tutorials. Usage pages focus on shared operations.

Model and Runtime Backend Selection

Model capabilities, runtime backends, and production topology belong in Model Support. Usage pages only explain how to pass these choices into the current tools.

API Schema

The recommended getting-started path currently focuses on WebUI and CLI. Detailed API fields, events, and asset formats will be organized in reference materials.

Next Steps