Earnote

Product Introduction

1. Definition: Earnote is a free, open-source (MIT license) desktop and mobile application for macOS, iOS, and iPadOS that functions as an AI-powered, privacy-focused lecture and meeting recorder. It is technically categorized as an offline-first AI transcription and note-generation tool.
2. Core Value Proposition: It exists to solve the problem of passive recording versus active note-taking by automating the creation of structured, actionable notes from audio recordings entirely on-device. Its primary value is delivering AI-powered note-taking without compromising user privacy, requiring subscriptions, or uploading sensitive audio data to the cloud.

Main Features

1. Local, On-Device AI Processing: Earnote's core architecture is built around local speech recognition and language model inference. It uses OpenAI's Whisper large-v3-turbo model for transcription and runs a local large language model (LLM) for note generation, both downloaded once and executed directly on the user's Apple Silicon Mac, iPhone 15 Pro, or later. This ensures zero data transmission for core functionality.
2. Structured Note Generation: Beyond raw transcription, the app's AI analyzes the transcript to produce a structured note containing a summary, key topics, actionable tasks (with checkboxes), and a list of generated flashcards. This transforms a linear audio recording into a hierarchical, study-ready document.
3. Integrated Study and Productivity Tools: Earnote is not just a recorder; it includes features for knowledge retention and workflow integration. It automatically creates flashcards (exportable to Anki) and a printable PDF study sheet. It can export notes and tasks to tools like Obsidian, Notion, Apple Notes, Things, and Todoist, and can name recordings based on calendar events.

Problems Solved

1. Pain Point: The cognitive load divide between listening/comprehending content (in lectures, meetings, or calls) and simultaneously writing coherent, structured notes. This often leads to incomplete notes, missed details, or poor retention.
2. Target Audience: University students (especially in lecture-heavy STEM and humanities courses), professionals in meetings and client calls, researchers conducting interviews, and any lifelong learner who attends talks or workshops. It specifically appeals to privacy-conscious users and those in environments with poor or restricted internet connectivity.
3. Use Cases: Recording a university lecture on Linear Algebra and receiving a summarized note with flashcards for exam preparation. Capturing a remote team meeting on Zoom (via system audio capture) and getting a concise summary with delegated tasks. Recording a personal explanation or study session to create a structured reference document.

Unique Advantages

1. Strengths & Limitations (Pros & Cons):

   • Pros: Unmatched privacy and data sovereignty with full on-device processing. No account, subscription, or ongoing costs (free and open source; one-time Pro purchase on iOS/iPadOS only). High utility density combining recording, transcription, note-taking, flashcard creation, and export in one tool. Transparent operation showing live transcription and note generation.
   • Cons: Platform lock-in to Apple ecosystem (macOS, iOS, iPadOS); no Windows or Android version. Hardware requirements are specific (macOS 15 on Apple Silicon, 8GB RAM; iPhone 15 Pro or later for full on-device iOS AI). Initial model downloads are large (~4GB total). Processing speed, while fast on Apple Silicon, is hardware-dependent and slower in Low Power Mode.

2. Key Alternatives & Differentiation:

   • Otter.ai / Descript: These are cloud-based, subscription services. Earnote differentiates by being a one-time purchase/free, offline tool, eliminating recurring costs and privacy concerns associated with uploading audio to third-party servers.
   • Apple Notes Voice Memo / Standard Recorder Apps: These provide basic recording and sometimes cloud transcription. Earnote differentiates by automatically generating structured notes, tasks, and study aids from the audio, moving far beyond simple audio playback or raw text transcription.
   • Manual Note-Taking (Notion, Obsidian): While flexible, they require full manual effort. Earnote acts as an automated input pipeline for these tools, generating the initial structured content that users can then refine and integrate into their existing PKM (Personal Knowledge Management) workflows.

Frequently Asked Questions (FAQ)

1. Is Earnote completely free and private? Yes, the macOS version is entirely free, open-source (MIT license), and processes everything locally. No audio or transcripts are ever uploaded to Earnote or third-party servers. The iOS/iPadOS app is free with an optional one-time Pro upgrade.
2. Can Earnote transcribe online meetings from Zoom or Microsoft Teams? Yes. Earnote can capture system audio simultaneously with the microphone, meaning it records all audio output from your computer, seamlessly transcribing all participants in a video call.
3. What are the system requirements for using Earnote on Mac? You need a Mac with Apple Silicon (M1 or newer) running macOS 15 (Sequoia) or later and at least 8 GB of RAM. An initial download of speech and AI models (~4GB total) is required.
4. How does the iPhone version work without a Mac? On iPhone 15 Pro and later models, the note generation AI runs directly on the device. For older compatible iPhones (iPhone 11+), the recording can be sent securely via iCloud to a linked Mac for processing, or you can use a cloud AI service with your own API key.
5. Why isn't the Mac app available on the App Store? The App Store's sandbox security restrictions would prevent Earnote from accessing system-level audio for recording calls and from writing files directly to other applications (like Obsidian vaults). The direct download is signed and notarized by Apple for security.