The text entry point voices the script itself, so there is no separate text-to-speech step to wire up before the lipsync runs. Because an avatar can be created from one image or one video at call time, the distance between a written script and a finished clip on disk is a single call.
A lipsync video generator with two entry points: supply audio, or supply text the service voices for you. The avatar is either an ID you already hold or one created on the spot from an image or a video.
- A video of a digital human avatar animated in sync with an audio file or audio URL you supply
- The same from text, with the audio generated first in a voice you name
- An avatar created from a video or an image, passed either as a URL or as a local file, when you do not already have an avatar ID
- Output written to a path you choose under a title you give it
- An async mode for the calls you do not want to sit and wait on
A Flyworks account with its token in FLYWORKS_API_TOKEN, and uv on your PATH. FLYWORKS_API_BASE_URL and FLYWORKS_MCP_BASE_PATH set the endpoint and base path the server works from.
One command plus a key — npx -y @smithery/cli install @Flyworks-AI/flyworks-mcp --client claude, then supply credentials