Facebook’s parent company, Meta Platforms, has launched a separate app for its Meta AI assistant. The app will offer generative AI-backed services similar to those offered by Google’s Gemini and OpenAI’s ChatGPT.
Prior to now, Meta AI was only available within the company’s bouquet of platforms – WhatsApp, Instagram, Facebook and Messenger. Meta said in a blog post that creating the standalone app is a first step toward building a more personal AI.
The separate app is powered by Llama 4, the company’s latest large language model. It is designed to rival the latest AI models from OpenAI, Google, Deepseek and Anthropic, with improved reasoning, multilingual capabilities and efficiency.
According to Meta, the chatbot will remember details, for example, the user’s travel interests or preferred language learning. If a user’s account is linked to their Facebook and Instagram accounts through the Accounts Centre, the updated AI can draw from both to enhance its understanding of users’ preferences.
Meta also announced that this is the app’s first version. Updates will follow soon:
“People can choose to experience a personal AI designed around voice conversations with Meta AI inside a standalone app. This release is the first version, and we’re excited to get this in people’s hands and gather their feedback,” the company announced.
The tech giant also indicated that the new app will integrate with Meta’s AI glasses and merge with the existing companion app.
A fascinating voice feature
The Meta AI features a voice demo that allows more natural and dynamic voice interactions. It is built with full-duplex speech technology. This means the AI can respond in real-time without pausing, providing more human-like conversation.
“This technology will deliver a more natural voice experience trained on conversational dialogue, so the AI is generating voice directly instead of reading written responses. It doesn’t have access to the web or real-time information”, the company describes its.
Also, it is designed for intuitive voice interaction. The “Ready to Talk” feature lets users enable voice by default, so starting a conversation is as easy as speaking, even while the user is multitasking on other apps.
Meanwhile, the company admits that the feature is experimental and may have some technical issues or inconsistencies, which will improve with time. Currently, the voice interactions and the full duplex demo features are available in the US, Canada, Australia, and New Zealand.
A personalised Meta AI app
Meta said that it has improved its AI to be more powerful and visually rich on the web. These include voice interactions, discover feed, enhanced image generation tools (style, mood, lighting, and colour presets), new document editor (currently being tested in select countries).
It will also include a discover feed, to allow users to share creative content that they have generated using the AI. The feed will look like any other known Meta apps (Instagram, Facebook). On it, users can like, share, comment and remix prompts.
Plus, it comes with a guarantee of control: “And as always, you’re in control: nothing is shared to your feed unless you choose to post it'”, Meta promised.
One interesting update is Meta’s promise that its AI will become more personal. “You can tell our AI to remember certain things about you (like that you love to travel and learn new languages), and it can also pick up important details based on context”, it announced.
Meta AI on the web is also getting an upgrade. It comes with voice interactions and the new Discover feed, just as will be seen on the app. This continuity across the app, AI glasses and the web helps deliver a more personal AI.
Personalised responses are available today in the US and Canada.
Technext Newsletter
Get the best of Africa’s daily tech to your inbox – first thing every morning.
Join the community now!
