Meta has launched the Meta AI app, a standalone app powered by Llama 4 that provides a personalized AI assistant. Integrated with Meta’s ecosystem including WhatsApp, Instagram, Facebook, Messenger, Ray-Ban Meta glasses, and the meta.ai website the app enables seamless interactions via voice and text. This initial release aims to enhance daily tasks and gather user feedback for future improvements.

A Personalized Assistant

The Meta AI app adapts to user preferences by drawing on information shared on Meta platforms, such as profiles and engagement history. For users in the US and Canada who connect their Facebook and Instagram accounts through the Accounts Center, the app combines insights from both to offer more tailored responses. For example, if you share an interest in travel, Meta AI can prioritize related suggestions. Users can specify preferences, like hobbies, to further customize the experience.

The app supports multitasking, allowing voice interactions while using other device features. A visible microphone icon indicates when the voice function is active. Additionally, the app offers image generation and editing, accessible via voice or text inputs, for creative and practical tasks.

Conversational Voice Features

Voice interaction is a key component of the Meta AI app, designed to feel intuitive. Powered by Llama 4, the assistant provides responses with a conversational tone. In the US, Canada, Australia, and New Zealand, a full-duplex speech demo is available, which can be toggled on or off. This experimental feature, trained on conversational dialogue, generates direct voice responses rather than reading text aloud. For now, it does not access the web or real-time data and may have technical inconsistencies, but Meta is collecting feedback to improve it. Visit the help center for details on managing voice settings.

Discover Feed for Shared Experiences

The Meta AI app includes a Discover feed, where users can explore and share AI interactions. The feed showcases popular prompts, which users can remix for their own use. Content is only shared if users choose to post, ensuring control over contributions. This feature encourages community engagement through AI-driven content.

Cross-Device Integration

The app integrates with Meta’s broader AI ecosystem, including Ray-Ban Meta glasses and the meta.ai website. For glasses users, the Meta AI app merges with the Meta View app, transferring existing settings and media to a new Devices tab. Conversations started on glasses can be continued in the app or on the website via the history tab, but chats initiated in the app or on the website cannot yet be resumed on glasses. This connectivity supports a consistent experience across devices.

The meta.ai website mirrors the app’s features, including voice interactions and the Discover feed, and is optimized for larger screens. It includes image generation with customizable presets for style, mood, lighting, and colors. Meta is also testing a document editor in select countries, allowing users to create text-and-image documents and export them as PDFs. Document import and analysis features are similarly in testing, expanding potential uses for professional tasks.

User Control and Privacy

Meta prioritizes user control, offering settings to manage personalization and voice features. Users can toggle the “Ready to talk” option for hands-free voice activation and choose what information the assistant uses, such as opting out of certain data sharing. The help center provides guidance on adjusting these preferences to balance personalization with privacy.

A Foundation for Future Enhancements

The Meta AI app is a step toward a more integrated AI experience, connecting Meta’s platforms and devices. By combining personalization, voice technology, and cross-device functionality, it aims to support everyday tasks. Meta plans to refine the app based on user feedback, with updates to enhance its capabilities.

Try the Meta AI app today on iOS or Android