The voice layer for agents
The Hook
“(No spoken dialogue — upbeat percussive synth beat)”
Product Reveal
“Hi, I'm Gradium's newest model, and I'm the fastest text-to-speech we've ever built.”
Feature Teaser
“Faster first audio, more natural conversations, no waiting, and guaranteed speed.”
Call to Action
“Try it today.”
Call to Action
“(No spoken dialogue — gentle synth outro)”
Synthesizable in Remotion · key components of the live-tts-latency-race template.
Horizontal multi-track audio playback waveform expanding along a time ruler with responsive millisecond readout badge.
Map audio buffer amplitude to SVG path height using useCurrentFrame and interpolate. Trigger horizontal mask translation tied to time-to-first-audio latency.
Box-and-whisker statistical markers rising smoothly from 0ms line with quartile range bars and floating numerical tags.
Use spring animations with staggered delay across items to interpolate box height, median notch, and whisker bounds against a calibrated vertical linear scale.
Purpose. Gamify latency comparison by initiating a race test.
Execution. Clicking pink CTA button triggers screen wipe into benchmark tracks.
Purpose. Demonstrate tangible performance lead using actual generated speech.
Execution. Gradium waveform finishes seconds ahead of stacked competitors.
Purpose. Reinforce live demo with aggregate benchmark distribution data.
Execution. Animated box plots demonstrate low median and minimal variance.
Purpose. Prompt immediate trial and brand recall.
Execution. Crisp centered CTA followed by minimalist brand icon animation.