Meta AI Launches Native Mac Desktop App With Screen Context and System-Wide Dictation
Meta has released a native Mac desktop app for Meta AI, built with AppKit and SwiftUI rather than Electron, weighing just 16MB. The 1.0 beta adds Mac-specific features including a global composer shortcut, system-wide dictation, and window-attachment for screen context.
Meta has released a dedicated Mac desktop app for Meta AI, marking a shift from web and mobile-only access to a native macOS client built specifically for Apple silicon.
The app launches as version 1.0 beta and installs at just 16MB. According to 9to5Mac's inspection, it uses an AppKit and SwiftUI shell with WebKit rendering for chat content — not an Electron wrapper or a repackaged iPad app. It requires macOS 15 or later and runs natively on Apple silicon Macs.
Mac-specific features
The app includes several features built around macOS system hooks rather than a generic chat interface:
- Quick Invoke: An Option-Space shortcut summons a compact Meta AI composer over any active app.
- System-wide dictation: Holding a shortcut lets users speak text that gets typed directly into any application, including Mail, documents, and code editors.
- Window attachment: With Screen Recording and Accessibility permissions granted, Meta AI can read visible text from another window and capture a screenshot to use as context for the next question. Meta describes this as context gathering only — the app does not currently control other applications.
- Dock-icon hiding: Users can remove the app from the Dock and rely entirely on keyboard shortcuts to invoke it.
The sidebar also includes Media, Artifacts, scheduled tasks, conversation history, and an "About Me" personalization section. Users can switch between thinking modes, attach files, generate media, and set up recurring briefings or reminders — features that mirror Meta AI's mobile app but are packaged with deeper OS integration.
Part of a broader push
The Mac app extends a string of Meta AI releases over the past several months. In April, Meta replaced Llama with Muse Spark as the underlying model. May brought more natural voice conversations and live camera-based assistance. In July, Muse-powered image generation arrived across Meta AI, WhatsApp, and Instagram — though Meta pulled back one feature that generated images from public Instagram posts shortly after launch. Meta AI also expanded into Threads DMs for private conversations, and most recently, Meta shipped Muse Code, a terminal-based coding agent for macOS and Linux.
Pricing for Meta AI's Mac app was not disclosed in the source material; Meta AI has historically been offered free of charge across its consumer surfaces.
What this means
This is a distribution and interface change, not a new model release — Meta AI's underlying model (currently Muse Spark) is unchanged. The significance lies in platform strategy: a native, lightweight Mac client with system-level hooks (global shortcuts, dictation, screen context) puts Meta AI in more direct competition with ChatGPT's desktop app and Anthropic's Claude desktop client, both of which have pursued similar OS-level integration on macOS. The screen-context feature, while limited to reading rather than controlling windows, signals where Meta is likely headed next — closer to the computer-use style agents OpenAI and Anthropic have already previewed. For now, the app's small footprint and native build suggest Meta is prioritizing speed and integration over feature parity with its mobile app.
Related Articles
Meta launches Muse Gadgets, open-source firmware and Linux SDK for building hardware that connects to its Muse agent
Meta introduced Muse Gadgets on Friday, an open-source project with firmware and a Linux SDK for building hardware that connects to its Muse personal AI agent. Meta also built a first-party device, Muse Home Link, and says it is giving away 5,000 units to Muse subscribers.
Suno adds Speech beta: spoken text and matching background music generated as a single audio track
Suno has added Speech, a beta feature that generates spoken text and matching background music together as a single audio track. Suno says it tested the feature with a small group for a month. The company has not disclosed how the underlying model was trained.
ChatGPT adds virtual try-on and Favorites in global shopping update built on Images 2.5
OpenAI launched two shopping features in ChatGPT globally on Thursday: a virtual try-on tool for clothing and accessories, and a Favorites function for saving products. OpenAI says both use its newly launched ChatGPT Images 2.5 model. Pricing and plan availability were not disclosed in the source reporting.
OpenAI launches Dots agent on GPT-6 Astra, limited to $20/month Pro tier and above
OpenAI unveiled Dots at DevDay 2026, a personal agent powered by GPT-6 Astra, available for now only on its $20/month Pro plan and above. Meta's rival Muse agent is free for anyone with a Meta account. OpenAI is positioning Dots as a long-horizon, knowledge-work tool with stronger privacy controls.
Comments
Loading...