Privacy-first open-source AI voice recorder as a hardware plugin
Phones can often feel cumbersome when quick action is needed. This is where a sleek, accessible, and intelligent solution comes in—an open-source, AI-powered pocket recorder. It syncs recordings locally, provides summaries, and runs on low-power ESP32-S3 hardware with Bluetooth and WebSocket support. The device attaches to a phone via MagSafe or works as a standalone unit, effortlessly capturing conversations, meetings, and spontaneous ideas. The first prototypes were developed in a San Francisco garage, and the project is now being made open-source to let others build and customize it. Your…
In plain words
This open-source AI voice recorder is a compact hardware device that captures audio conversations and ideas on an ESP32-S3 chip. It attaches to phones via MagSafe or operates independently, syncing recordings locally and generating summaries through on-device AI. The device runs on low power and connects via Bluetooth and WebSocket. Built as a privacy-focused alternative to cloud-dependent recorders, it allows users to build and customize their own units from open-source designs.
written from the facts on this page · September 2026
From the sources
In the maker’s words, at launch
Phones can often feel cumbersome when quick action is needed. This is where a sleek, accessible, and intelligent solution comes in—an open-source, AI-powered pocket recorder. It syncs recordings locally, provides summaries, and runs on low-power ESP32-S3 hardware with Bluetooth and WebSocket support. The device attaches to a phone via MagSafe or works as a standalone unit, effortlessly capturing conversations, meetings, and spontaneous ideas. The first prototypes were developed in a San Francisco garage, and the project is now being made open-source to let others build and customize it. Your feedback on the device is highly valued! The landing page is live: https://www.openvision.engineering/, and you can watch the product video here: https://www.youtube.com/watch?v=XupW9pi11MI. If you’re in San Francisco, you’re welcome to see it in action! Located in Presidio, the creators are happy to meet in person and demonstrate the product.
More ai this month
the category →
I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.
AI · 17d ago · simedw.com
Astute▲585Automate your B2B brand going viral, with new media creators
AI · 18d ago · company-app.joinastute.com
Hey HN, Henry from Cactus here! We previously released Cactus Needle, a 14MB agentic LLM for tool call, device use, and structured extraction for phones, wearables, smart homes, small robots and microcontrollers. We got really great feedback here, and have now incorporated the suggestions to release Needle 2. The whole model is a single 14MB binary that runs a full session in 28MB of RAM; 45m parameters at 2bit compression. Needle hits 500 tokens/sec decode speed on a Raspberry Pi 5, sits between 400-1,500 tokens/sec on VR devices like Meta Quest 3S and Apple Vision Pro, and ranges…
AI · 27d ago · cactuscompute.com
