Make a speaking presenter
Portrait + spoken script → lipsynced introduction.
Muse generates a fictional museum guide while MiniMax Speech reads the workshop introduction. LongCat combines the portrait and speech into a video, using a delivery brief to control movement. The two branches let you change the presenter and the spoken words independently.
Saved result
A fictional museum guide reads the workshop introduction. Sampled frames retain the face; model-assisted audio review recovered the script. Watch playback to judge lip sync.
Make it yours
- Open the graph — save first if the canvas already has your work. Undo brings yours back.
- Change Presenter look, Spoken script and Delivery. Keep the face and mouth visible, match the selected voice to the presenter, and keep the script under 30 seconds. LongCat renders at 480p.
- Check the editor estimate, then run with your NanoGPT key. The saved workshop run reported at least $0.28 for generation; some provider price fields were missing. Its separate model-assisted audio review reported $0.00164. Longer speech increases video cost. Check the editor estimate before running.
- Play with sound. Check the spoken words, face consistency and lip-sync timing through the whole clip.
The current workflow uses the provider’s renamed avatar model ID. The original sampled graph is preserved below.
Download workflow · Original sampled graph · Explore MCP tools · awesome-noodles