GemmaNotes

Push-to-talk voice notes for Obsidian, transcribed entirely on-device with Google's Gemma 4 (E2B / E4B) running in-process via transformers.js. No server, no API key — fully offline after a one-time model download.

Built with Love and Antigravity CLI, Gemini 3.5 Flash

      ▄▀▀▄        Antigravity CLI 1.0.2
     ▀▀▀▀▀▀       xxxxxxxxxxxxxxxxxxxxx
    ▀▀▀▀▀▀▀▀      Gemini 3.5 Flash (High)
   ▄▀▀    ▀▀▄     ~/gemmanotes
  ▄▀▀      ▀▀▄

How it works

1. Toggle recording with the ribbon mic button or the Toggle voice recording command (bind a hotkey to it).
2. On stop, a placeholder is pinned into the current note.
3. Audio is decoded to 16 kHz mono, split into ≤25 s chunks, and transcribed by Gemma 4. Transcriptions run through a FIFO queue, so you can start the next recording while one is still processing.
4. The placeholder is swapped for the transcribed text.
5. A status-bar hint offers a one-click rewrite into tidy prose.

Local Setup

Git clone:

gh clone https://github.com/sarath/gemmanotes

Development:

npm install
npm run build      # production bundle -> main.js
npm run dev        # watch mode

Install into a vault for local testing:

Mac

git clone https://github.com/sarath/gemmanotes
./install.sh /path/to/your/vault <feature-branch>

This fetches origin, checks out origin/<feature-branch> cleanly, runs npm run build, and copies main.js, manifest.json, and styles.css into <vault>/.obsidian/plugins/gemmanotes/. Then enable the plugin and use Settings → GemmaNotes → Download to fetch the model (~3.2 GB for E2B, ~5 GB for E4B).

Local Testing and Development

If you are developing inside Google Cloud Shell and want to deploy, test, and debug the plugin directly in a local Obsidian instance:

1. Tunnel Remote Debugging Port

Start your local Obsidian instance with remote debugging enabled (e.g., obsidian --remote-debugging-port=9222). Then, run the following command to tunnel port 9222 from your local machine to your Cloud Shell workspace:

gcloud cloud-shell ssh --ssh-flag="-L 9222:localhost:9222"

This allows scripts running in the Cloud Shell environment to connect to the Chrome DevTools protocol on your local Obsidian instance.

2. Build, Deploy, and Test

1. Compile the production bundle:

   npm run build

2. Deploy the files (main.js, manifest.json, styles.css) to the connected Obsidian instance and automatically reload the plugin:

   npm run local:deploy

3. Test model loading and verify initialization:

   npm run local:test-load

Known v1 limitations

• Fixed-window chunking — long recordings are cut on a 25 s boundary, which can split a word. Silence-aware chunking is the planned next step.
• transformers.js Gemma 4 audio API — the processor/model call in src/transcriber.ts follows the documented image-text-to-text pattern; the exact audio surface should be confirmed against the model card sample code.
• Undo granularity — placeholder→text and rewrite swaps are written via the vault API; they are not always a single editor-undo step.
• Model cache — stored in the browser Cache API. Persistent and offline-safe, but not yet relocated to the plugin data dir.
• Desktop only — the model size and WebGPU requirement rule out mobile.