Perplexity AI has launched a new voice assistant feature in its iOS app, significantly enhancing the AI assistant's practicality and user experience. According to AIbase, the new feature supports various tasks such as setting alarms, finding directions, sending messages, and making restaurant reservations. Combined with powerful real-time search and multi-app integration, it offers users a seamless smart living experience. The update is now available on the App Store, with enthusiastic community feedback, marking Perplexity's strong foray into the comprehensive AI assistant field.
Core Functionality: Multitasking and Real-time Interaction
The Perplexity voice assistant significantly improves the automation of daily tasks through multimodal input and application integration. AIbase has summarized its main features:
Voice-driven task execution: Users can set alarms, send text messages, make phone calls, or manage calendars using voice commands, such as "Set an alarm for 7 am tomorrow" or "Send Sarah a meeting invitation."
Real-time route planning: Integrating with map services like Gaode Map, the voice assistant can find and plan routes based on instructions, such as "Find the fastest route to the nearest coffee shop," and provide real-time traffic updates.
Multi-app integration: Supports integration with media services like Spotify and YouTube Music to play music, podcasts, or videos; it can also book restaurants or hail rides through third-party apps.
Screen and camera interaction: Supports "Live View" camera queries and the "On-Screen Context" function, which can analyze screen content or real-world objects, such as translating road signs or summarizing web text.
Multilingual support: Added voice interaction in Japanese, Spanish, and other languages, combined with natural language processing, ensuring smooth cross-language conversations.
AIbase noted that in community tests, users completed the entire process of searching, filtering, and booking a restaurant for four people using the voice command "Find a restaurant tonight and book a table for four," with an intuitive and efficient overall experience comparable to Siri and Google Assistant.
Technical Architecture: Multimodal AI and Context Awareness
The Perplexity voice assistant is based on its core AI models (such as Claude3.7Sonnet, GPT-4o, Gemini2.5Pro) and a multimodal technology stack. AIbase analysis shows that its key technologies include:
Automatic Speech Recognition (ASR): Uses deep neural networks to accurately recognize multilingual voice input, supporting complex commands and accent parsing, with response latency as low as milliseconds.
Context-aware engine: By remembering conversation history, the assistant can seamlessly handle subsequent instructions, such as booking a restaurant after searching for one, without repeating the context.
Multimodal processing: Integrates vision (camera input), audio (voice commands), and text (screen content) to support cross-modal tasks, such as "Translate the French road sign in this picture."
Real-time search and citation: Combining Perplexity's search technology, it provides answers with sources to ensure information accuracy, such as providing a link to a weather website when querying "Barcelona weather today."
Security and privacy: Runs in a sandbox environment, data is encrypted during transmission, voice queries are still recorded in incognito mode, requiring manual clearing by the user.
Currently, the voice assistant is available on iOS (requires iOS 16.0+), while some features on the Android version require update 2.37.0. The Mac version is forthcoming. AIbase believes that its open-source API and multi-model selection provide developers with flexible customization options.
Application Scenarios: From Daily Convenience to Professional Assistance
The multi-functionality of the Perplexity voice assistant covers a wide range of needs, from personal life to work scenarios. AIbase summarizes its main applications:
Daily life management: Setting reminders, sending messages, playing media, or booking services simplifies daily tasks, such as "Remind me to watch the new Netflix show at 8 pm tonight."
Travel and navigation: Real-time route finding, translating road signs, or booking hotels is suitable for travelers and international users, such as "Find the bus route to the Tokyo Tower."
Improved work efficiency: Managing calendars, drafting emails, or summarizing notifications helps professionals work efficiently, such as "Summarize my unread emails and mark priorities."
Education and research: Use voice to query academic materials or analyze screen content, supporting students and researchers, such as "Summarize the key points of this PDF."
Accessibility support: Provides voice control and environmental awareness for visually impaired or mobility-impaired users, enhancing device accessibility.
Community feedback shows that the voice assistant's context memory and multilingual support are outstanding in cross-cultural communication, and it has been praised as a "Siri replacement for iPhone users." AIbase observes that its integration with Telegram Bot further expands cross-platform usage scenarios.
Getting Started: Easy Enablement, Quick Experience
AIbase understands that the Perplexity voice assistant is now open to all users through the iOS version of the Perplexity app; Android users need to update to version 2.37.0. Users can quickly get started by following these steps:
Update the Perplexity app from the App Store or Google Play (iOS 16.0+ or Android 10+);
Open the app, click the homepage banner, or go to settings to enable the voice assistant (Settings > Enable Assistant);
Grant necessary permissions (microphone, camera, location, contacts, etc.), activate the assistant through gestures (such as pressing the power button) or the interface;
Use voice commands, such as "Set an alarm for 9 am tomorrow" or "Find the route to Paris," and view the results in real time.
The community recommends enabling "Hands-Free Mode" for continuous conversation and clarifying instructions to optimize multitasking. AIbase reminds that Android users may need to sideload the 2.37.0 APK to get full functionality, and iOS users should pay attention to privacy settings to manage query history.
Community Feedback and Improvement Directions
After the release of the voice assistant, the community gave high praise to its multi-app integration and natural interaction. Developers called it a "perfect combination of search and task automation," especially surpassing Google Assistant in route planning and media playback. Japanese users particularly appreciated its UI feedback and smooth voice, providing a "sense of security." However, some users pointed out that the assistant lacks a hotword wake-up like "Hey Google," and queries are still recorded in incognito mode, which may raise privacy concerns. The community also expects support for more languages (such as Chinese) and video analysis functions. Perplexity responded that future updates will optimize hotword wake-up and enhance privacy controls. AIbase predicts that the assistant may integrate with the Comet browser or enterprise API to build a cross-device AI ecosystem.
Future Outlook: The Evolution of the Smart Assistant Ecosystem
The launch of the Perplexity voice assistant demonstrates its ambition to transform from a search tool to a comprehensive AI assistant. AIbase believes that the combination of its multimodal interaction and real-time search lays the foundation for challenging giants like Siri and ChatGPT. The community is already discussing integrating it with Home Assistant or the MCP protocol to build smart home and automated workflows. In the long term, Perplexity may launch an "AI assistant marketplace," providing customized voice models and third-party plugins, similar to the Alexa Skills ecosystem model. AIbase expects the full launch of the Android and Mac versions of the assistant in 2025, as well as breakthroughs in multimodal tasks and low-power device support.
