Google unveiled AI Edge Foresight, a meeting note-taking app designed to work entirely offline using advanced local AI models. The app aims to compete directly with Granola by offering transcription, real-time note generation, and an AI assistant, all while preserving user privacy through edge computing.
- Works fully offline using local EmbeddingGemma 2 AI model
- Offers split-screen manual and AI-generated meeting notes
- Supports document uploads for enhanced real-time Q&A
What happened
Google introduced AI Edge Foresight, an AI-powered meeting note-taking app for Mac that functions entirely offline. The app employs Google's on-device EmbeddingGemma 2 model, which contains 740 million parameters, enabling transcription, note generation, and real-time assistance during meetings without needing internet access.
The app features a split-screen interface similar to the popular AI note-taker Granola, allowing users to manually jot notes on one side while AI-generated summaries appear on the other. Additionally, users can review full transcripts of meetings and interact with the Gemma 4-powered assistant for contextual answers during discussions.
Why it matters
AI Edge Foresight's full offline functionality addresses privacy and connectivity challenges faced by professionals who take notes in sensitive or low-network environments. By processing data locally, Google provides users a secure solution while showcasing the capabilities of its advanced Gemma series AI models.
The app also stands out by integrating a personal knowledge base using uploaded documents in various formats, including PDFs and Google Docs, allowing for dynamic, data-informed responses during meetings. This feature positions Google to compete in a crowded market where numerous startups and established players are launching AI-powered note-taking tools.
What to watch next
It will be important to see whether Google plans to expand AI Edge Foresight’s availability beyond Mac and offline use to integrate cloud-based AI services across video conferencing apps. Such an expansion could strengthen its competitive edge against established note-taking solutions like Granola and newcomers from companies like Wispr and Calendly.
Users and industry observers should also monitor how Google further enhances its local AI models and the user experience, as well as how it navigates privacy, accuracy, and interoperability challenges in real-world meeting environments. The evolving productivity software landscape suggests Google’s offline AI approach could usher in new standards for privacy-centric work tools.