Talaat’s AI-powered meeting notes stay on your device, not in the cloud

AI note-taking app Granola, valued at $250 million, has become a popular tool among tech industry founders and venture capitalists. But one developer believes there’s demand for a more private, local-only alternative that’s available for a one-time fee and without a subscription. This led to the creation of a new Mac app called Talat.

Nick Payne, a developer based in Yorkshire, England, and a self-described computer geek, says the idea to create a homegrown AI blogging machine came mostly because of a series of happy accidents.

“I think Granola is great; it’s a shining example of what you can do with the Electron app [a framework for building desktop applications] “I was given enough love and care,” he told TechCrunch. “When I first tried it, I was impressed that it was able to record system audio on my Mac without recording video, which was the standard workaround at the time. This led to me doing a lot of research, and discovering a relatively new and poorly documented Apple API.

To make it easier to work with the Core Audio Taps API, which lets developers tap into Mac audio streams, Payne decided to create an open source audio library, AudioTee.

“During that time, I was slowly putting together a kit, but I never found anything that looked like it could stand on its own as a product rather than just a cool tech demo,” Payne said. “The modern hosted transcription models — the same providers like Granola use — are amazing, and it’s pretty cool to see your speech appear on the screen in near real-time. But it’s always bothered me that the trade-off requires providing not just my data, but my audio data; my actual voice,” he added.

He then found a software toolkit called FluidAudio, which is a Swift framework that enables fully local, low-latency audio AI on Apple devices. It lets you run small, fast copy models directly on your Mac’s Neural Engine, Apple’s dedicated AI processing machine.

That was the piece that made Payne realize he could turn his research into an actual product — one where your voice never leaves your Mac and your texts aren’t stored on another company’s servers.

Talaat, built alongside Payne’s old friend and former bandmate Mike Franklin, is the result of Payne’s interest in the audio field. The result is a one-time, 20MB purchase that doesn’t require you to create an account or even share analytics data again with developers. There are no ongoing fees either.

While some AI note-taking tools may have more features, Talat offers a simplified set of features. It captures audio from your computer’s microphone when you’re in meeting apps like Zoom, Teams, Meet, etc., and transcribes it in real-time. The app attempts to assign speakers in real time, but you can reassign them as needed. You can also take notes, as well as edit, delete, or split text sections. When the meeting ends, the local LLM creates a summary that includes key points, decisions and action items.

Notes, texts and summaries can be searched in Talat as well.

In addition to the privacy angle, Payne said the goal is to give users more options.

“We lean into configurability and allow users to control where their data goes: choose your MBA, auto-export to… [notetaking app] Obsidian, webhooks that push data out when a meeting ends, and an MCP server, a standardized way for AI tools to connect to external data sources, “to pull it on demand,” he explained.

Under the hood, the AI ​​is a patchwork — “mostly cobbled together and abstracted behind FluidAudio,” noted Payne, who credits it with doing a lot of the heavy lifting. For the recap piece, the app defaults to an Al model called Qwen3-4B-4bit, which can run even on fairly modest hardware.

However, users can choose to route this to any LLM cloud provider of their choice, or they can choose between two variants of Parakeet – speech recognition models developed by Nvidia – or point it to Ollama (a tool for running AI models locally), giving them more control over the experience. Over time, Talat will add support for more built-in selections and will have integrations with other apps, such as Google Calendar and Notion.

At launch, users of M-series Mac computers (those using Apple’s own processors, starting with the M1) can download the app and try it out for free with 10 hours of recordings before making a purchase.

Talat is available for $49 in this beta version, which is still in active development.

When the app reaches version 1.0, the price will increase to $99.

Payne and Franklin are bootstrapping Talaat and plan to keep the product basic with a one-time purchase from now on.