What ElevenLabs shipped
ElevenLabs brings voice-driven workflows to iOS and Android with the launch of Dictation on their mobile app. Powered by their Scribe v2 speech-to-text model, the feature lets creators speak their prompts to generate voiceovers, images, or video on the go.
At 28 seconds, the announcement film is built for the same platforms it targets—TikTok, Reels, and Shorts. It relies on rapid-fire UI sequences and sharp typography to communicate speed and accuracy.
How the motion works
The edit prioritizes momentum, using kinetic-typography to mirror the act of speaking. Text scales and snaps into place synchronously with the voiceover, establishing a direct visual link between the user's voice and the system's transcription speed.
UI interactions are stylized but legible. Instead of a slow screencast, the film uses a punch-in to isolate the microphone button, immediately establishing the core interaction model. A subtle camera-drift during text generation keeps the frame alive while the AI processes the prompt.
Transitions rely on the hard-cut between the dictation interface and the generated outputs—audio waveforms, images, and video. This structural choice compresses time, making the journey from spoken prompt to final asset feel instantaneous by intentionally omitting the actual processing wait times.
What to steal from it
- Use kinetic typography to visualize audio inputs when demonstrating voice-first features.
- Punch in on specific UI elements to establish interaction models without showing the entire screen.
- Compress time with hard cuts from input to output, bypassing realistic but tedious loading states.
- Keep the camera drifting slightly during static UI moments to maintain visual momentum.
Want a video like this for your product?
Impractical cuts one from a single prompt — the same motion craft, in about twenty minutes.
