Submit iOS meeting notes to Obsidian using local AI transcription

This workflow enables seamless, privacy-first capture of meeting notes from your iPhone. It pairs with an iOS Shortcut that leverages Apple’s on-device transcription from the Voice Memos app and optionally passes the output to ChatGPT or a local AI model for summarization.

Who it's for

Anyone who wants fast, secure note capture on iOS Professionals (e.g., lawyers, therapists) who require on-device processing for privacy Obsidian users who want to sync mobile notes via Google Drive

What it does

You record a voice memo in the iOS Voice Memos app. The Shortcut transcribes it locally (no API or cloud involved). Optionally, a summarization step is done via ChatGPT or a replaceable local model. The data is sent to an n8n webhook, where it’s converted into a .md file. The Markdown file is uploaded to a Google Drive folder synced with your Obsidian vault.

Key Benefits

🧠 Keeps your meeting notes private — no cloud APIs required 🗃️ Easily searchable in Obsidian as structured Markdown files 🛠️ Fully local if you swap out ChatGPT for a local model (can be less stable)

Limitations

Transcriptions longer than ~1 hour may fail or produce unstable results. Some setup required to replace ChatGPT with a local model in the Shortcut.

Setup

Install and configure the iOS Shortcut Replace the ChatGPT step in the Shortcut if you need full local-only processing Point the webhook in the Shortcut to your n8n instance Make sure your Obsidian vault is synced with the Google Drive folder used in the workflow Update any of the prompts in the iOS shortcut to ensure its personalized.

> ⚡ Pro Tip: Use a Set node early on to clearly define the filename, title, and text so it's easier to modify this for other note types (e.g., journals, therapy sessions, etc.)

0
Downloads
0
Views
6.83
Quality Score
beginner
Complexity
Author:Angel Menendez(View Original →)
Created:11/26/2025
Updated:12/2/2025

🔒 Please log in to import templates to n8n and favorite templates

Workflow Visualization

Loading...

Preparing workflow renderer

Comments (0)

Login to post comments