Google quietly launched an AI dictation app that works offline

Google has quietly rolled out a new AI-powered dictation app for iOS that processes speech entirely on-device, putting it in direct competition with established tools like Wispr Flow. The release received little fanfare, but the technology behind it signals a meaningful shift in how offline AI tools are being deployed.

The Details

The app is built on Gemma, Google's family of lightweight AI models designed to run locally without requiring a cloud connection. This offline-first architecture means user audio never leaves the device, which has clear privacy implications.

Key points about the release:

• The app runs fully on-device, with no data sent to external servers during transcription
• It uses Gemma AI models, the same open-weight model family Google has been positioning for edge and local deployments
• The launch was low-key, with no major announcement or press push from Google
• It enters a market currently occupied by tools like Wispr Flow, which has built a following among professionals who rely on fast, accurate voice-to-text

The quiet nature of the launch suggests Google may still be refining the product, but the fact that it is publicly available on iOS indicates it is far enough along to compete in the real world.

Why This Matters for Service Providers

For MSPs and telecom resellers advising clients on productivity tools, offline AI capabilities are becoming a critical evaluation criterion, particularly for clients in healthcare, legal, finance, and government sectors where data sovereignty matters.

A Google-backed offline dictation tool adds legitimacy to the on-device AI category and could accelerate client demand for similar privacy-preserving solutions across other workflows. If your customers are currently paying for cloud-dependent transcription tools, this is a good time to revisit those conversations.

The broader signal here is that AI is moving to the edge. Vendors are no longer treating cloud connectivity as a requirement, and service providers who understand this shift will be better positioned to guide clients through the next wave of AI tool adoption.

This also reflects growing competitive pressure in the voice AI space overall. As more players, including major platforms like Google, enter the offline and on-device market, the bar for what clients expect from AI voice tools will continue to rise.

Looking Ahead

Watch for Google to expand the app's capabilities or bring similar on-device AI features into its broader Workspace ecosystem, which would have direct implications for MSPs managing Google-based environments. If your clients are heavy voice-to-text users, it is worth testing this tool now before they find it on their own.

For the full story, read the original article on TechCrunch AI.