A Seedance render from our Prompt Library, shown as an example.

/talking-ugc🎬 AI VideoVeo 3.1Native sound114 creditsNew

Talking UGC with audio in Claude & ChatGPT

A creator talks to camera with lip-synced dialogue and room sound, in one take.

Example prompt

@ClipSpeed Make an 8-second selfie video of a creator at the gym saying “I finally found a pre-workout that doesn’t taste like chalk.”

Start in 3 steps

  1. Add ClipSpeed to your AIOne URL for everything: https://api.clipspeed.ai/mcp. Sign in with ClipSpeed, no API key.Add to ClaudeOther apps
  2. Prepare your briefThe exact line you want said (one or two sentences), who says it and where.
  3. Run your first promptCopy the example, open a new chat, attach anything it needs and send. You see the credits before anything runs and say yes in the chat.

How it goes in the chat

@ClipSpeed Make an 8-second selfie video of a creator at the gym saying “I finally found a pre-workout that doesn’t taste like chalk.”

Veo 3.1 makes the dialogue and the room sound together. I’ll keep the line short so it fits.

ClipSpeedPrompt for this example
He holds the phone out at arm's length, a little to one side, his shoulder in the near corner, framed chest-up. The phone in his hand is the one recording this, and he talks straight into its lens. Early twenties, dark tight curls on top with short sides, light stubble, fitted charcoal tank. He is mid-sentence and animated, a big open grin breaking through as he speaks, eyebrows lifting, head bobbing with the words, the frame drifting slightly as he moves. Bright commercial gym behind him, a drop ceiling with fluorescent panels, dumbbell racks, flat benches, cable machines, people training, a wall of windows onto green fields, daylight. Natural skin texture, a light post-workout sheen. One unbroken handheld take, no on-screen text, no logos, no captions.
9:16Veo 3.1With audio114 credits

This costs 114 credits. Make it?

114 credits

Yes, make it

Done. Lip-synced line with gym sound.