Perplexity Expands Voice Interaction in Comet Browser

Perplexity announced a significant upgrade to the voice capabilities of its Comet browser, moving the experience beyond a simple query tool to full‑session, hands‑free navigation. Desktop users can activate the new mode immediately, while iOS users will receive the functionality a few days later, according to pre‑order listings.

The upgraded voice mode runs on OpenAI’s latest real‑time model, identified as gpt-realtime-1.5. This model is designed for low‑latency voice agents, delivering quicker responses and more natural pacing that supports longer listening sessions. Perplexity’s CEO, Aravind Srinivas, highlighted the rollout on the social platform X, thanking OpenAI’s multimodal team for their collaboration.

Unlike most browsers that treat voice as a “party trick” limited to search queries, Comet now lets users control virtually every browsing action with spoken commands. Users can open websites, scroll pages, click links, and even request summaries or comparisons of content—all without a mouse or keyboard. Sample commands include “scroll down,” “open the third link,” “summarize this page,” and “compare it to the tab on the left.” The activation shortcut on Windows is Shift + Alt + V, and on macOS it is Shift + Option + V. A comparable shortcut will be introduced for iPhone when the iOS version launches.

Perplexity reports that the new voice mode improves tool‑call stability by more than 25 percent, reducing misfires when the browser attempts to execute spoken instructions. The voices themselves have been refined for clearer articulation and smoother pacing, enhancing usability for extended interactions.

Privacy is a core component of the rollout. The browser processes voice data locally whenever possible and does not store click histories in the cloud by default, aiming to avoid the creation of ad‑tracking profiles. This stance differentiates Comet from many competitors that rely on server‑side processing and extensive data collection.

Looking ahead, Perplexity plans to enrich the voice experience with a learning assistant that adapts to individual user preferences. The assistant could help with tasks such as shopping, ordering food, or finding flights based on past behavior. Additional features under development include a password manager and cross‑device synchronization, though Android users will need to wait for those capabilities.

The launch reflects a broader industry shift toward ambient computing, where interaction is increasingly voice‑driven and context‑aware. By offering a robust, privacy‑focused, and hands‑free browsing experience, Perplexity positions Comet as a forward‑looking tool for users who prefer to navigate the web through conversation rather than clicks.

Used: News Factory APP - news discovery and automation - ChatGPT for Business