23 Daily — According to github.com, Yogurt is a free and open-source desktop meeting assistant that gained attention on November 11, 2026. It uses on-device AI models to transcribe, summarize, and enable Q&A for meetings without uploading any data to the cloud, emphasizing privacy and data security.

Local AI Technology Enables Private Meeting Assistant

Yogurt runs in the background without joining calls as a bot, using built-in local AI models to process recording data. It supports macOS 14.4 (Apple Silicon), Windows 10 version 2004 or later, and Ubuntu 22.04 or similar (x64 AppImage), requiring at least 8GB memory. This design effectively prevents meeting content from leaving the user's computer, ensuring privacy protection.

Features Focused on Meeting Transcription, Summarization, and Q&A

Yogurt offers real-time meeting transcription and automatically generates summaries to help users quickly grasp key points. Additionally, it supports interactive question-answering based on meeting content to enhance efficiency. It also provides a local MCP server, allowing users to optionally share data directly with advanced models like Anthropic or OpenAI, avoiding paid cloud intermediaries and giving users control over data flow.

Differences and Improvements Over Existing Open-Source Tools

Creator Eric shared that after trying many popular open-source note-taking tools, common issues included lack of calendar integration, absence of a local MCP server, or mandatory account creation. In contrast, Yogurt focuses on comprehensive functionality and user experience by fine-tuning UI interactions and animations, optimizing AI model calls for performance, making the software smoother and more user-friendly.

Editor's take: Local AI Assistants as a New Entrepreneurial Opportunity

Yogurt reflects a trend of leveraging smaller local AI models to replace dependency on large cloud-based models, appealing to privacy- and cost-conscious users. Entrepreneurs can see that combining local AI technology with open-source approaches to offer a high-efficiency meeting assistant without relying on the cloud presents a market opportunity. Verifying the product's value can involve deploying it across various OS environments to test transcription accuracy, summary quality, UI responsiveness, and surveying user acceptance of privacy and offline operation.