Voice Command for PS5, currently in preview for eligible US and UK accounts in English, enables eligible players to use natural speech to find or open games, apps, and settings. They can also control supported media and access help resources.

Together, these capabilities show how voice can become more than an input method. It can reduce friction, support accessibility, and keep players immersed.

However, this was more than a straightforward API integration. The SIE team evolved the speech platform from a single vendor integration into a configurable, multi-provider architecture and designed the platform architecture and capabilities needed to integrate and operate speech technologies reliably across the PlayStation platform experiences. The work included building common approaches for evaluation, routing, observability, and adaptation so the underlying technology could evolve without requiring each user experience to be redesigned.

As Golda James, Director of Engineering at Sony Interactive Entertainment explains, balancing these requirements at scale was a major engineering hurdle. “At this scale, this is not an easy problem to solve. We had to balance accuracy, latency, and cost—the trifecta. We established a solution that could scale globally, transcribe in real time, and domain adaptation for game titles and specialized gaming terminology.”

Cloud Speech-to-Text converts spoken audio into text, while Cloud Text-to-Speech generates spoken output for supported experiences. Adaptation capabilities help improve recognition of specialized gaming vocabulary, making transcripts and voice interactions more relevant to the way players communicate.

SIE engineers designed the integration so relevant game titles, character names, abbreviations, numbers, and other specialized terms can be prioritized as gaming terminology evolves. This approach helps recognition stay aligned with the language that players use without requiring each experience to change whenever vocabulary is updated.

SIE also established a repeatable evaluation approach to assess speech services and model updates across accuracy and transcription quality, using representative inputs spanning languages, regions, and gaming terminology. The team evaluates end-to-end latency, streaming responsiveness, reliability, regional performance, scalability, and cost. The result is a speech platform designed to support high-volume, real-time interactions while remaining adaptable to new games, terminology, languages, and player needs.

For SIE, a flexible speech foundation provides room to evolve alongside advances in speech technology while continuing to focus on the factors that matter most to players: quality, responsiveness, reliability, and accessibility.

For players, the value is straightforward: fewer interruptions between intent and action.

Across PlayStation experiences, chat transcription makes conversations easier to follow, voice dictation provides an alternative to typing, broadcast chat-to-speech helps creators stay focused on gameplay, and Voice Command offers eligible players a conversational way to navigate supported console experiences.

The work established a scalable speech platform designed and operated by SIE engineers to adapt to the language of games, support voice experience at scale, and evolve alongside advances in speech technology.

By combining SIE’s understanding of players with Google Cloud’s speech capabilities, SIE is building PlayStation voice experiences designed to feel faster, more natural, and less disruptive, helping players spend less time navigating interfaces and more time immersed in play.