Figma Launches 'Voice Tokens' to Standardize Accessible Voice Interfaces in Design Systems

Tech · 5 min read

Figma's new Voice Tokens allow teams to define vocal attributes — such as speaking rate, verbosity, confirmation patterns, and error phrasing — as first-class tokens in a shared design system. Tokens can be applied to components, letting designers and engineers export consistent voice behavior for screen readers, voice assistants, and in-app narration.

The feature includes presets informed by WCAG and cognitive accessibility research: for example, a 'calm-guidance' token that lowers speaking rate and avoids filler words, and a 'concise-confirmation' token for critical flows like payments. Teams can also attach context rules so tokens switch automatically depending on user preferences or device capabilities.

Integrations with popular prototyping and dev tools mean voice token values can feed into runtime libraries and automated tests. Design leaders say the tokens reduce ambiguity between design intent and engineering implementation, making inclusive voice interactions repeatable across platforms.