Usage

Commands

All commands are available via the command palette (Ctrl/Cmd+P).

Command	Description
Transcribe current image	Transcribes the currently active image file. Only appears in the palette when an image file is open.
Transcribe all images in current note	Finds all embedded images in the active note and batch-transcribes them. Only appears when a Markdown file is open.
Transcribe images in folder…	Opens a folder picker, then lets you select which images to transcribe via a checkbox list.
Install AI model	Opens a picker listing recommended models. You can also type any Ollama model name to install it (e.g. `llava:13b`).
Select AI model	Opens a picker to choose which installed model to use for transcription. The current model is marked.
Remove AI model	Opens a picker to delete an installed model from Ollama and free disk space.

File explorer

Right-click in the file explorer to access transcription actions:

On an image file: “Transcribe image” — transcribes that single image
On a folder: “Transcribe all images in folder” — batch-transcribes all images in the folder

Editor

Right-click on an image embed in the editor (e.g. ![[photo.png]] or ![](photo.png)) to see:

Transcribe this image — transcribes the embedded image

Model Management

You can manage AI models entirely from the command palette — no terminal needed.

Installing a model

Open the command palette and run Install AI model
Select a recommended model from the list, or type any Ollama model name (e.g. llava:13b, qwen3.5:9b-q4_K_M)
A progress notice shows download status with percentage
Once installed, the model is automatically selected for transcription

You can also install models from Settings > Transcriber using the recommended models list or the custom model field. See Tips for help choosing a model.

Selecting a model

Open the command palette and run Select AI model
The picker shows all models installed in Ollama
The currently selected model is marked
Pick a model to switch to it immediately

Removing a model

Open the command palette and run Remove AI model
The picker shows all models installed in Ollama
Select a model to delete it from Ollama and free disk space

Transcribing a Single Image

Right-click an image in the file explorer and select Transcribe image, or open an image and use the command palette (Transcribe current image)
A notice appears while transcription is in progress
When complete, a .md file is created alongside the image with the same name (e.g. photo.png produces photo.md)

Transcribing All Images in a Note

Open a Markdown note that embeds images (via ![[image.png]] or ![](image.png))
Open the command palette and run Transcribe all images in current note
All embedded images are detected, deduplicated, and batch-transcribed
Progress and summary notices work the same as folder transcription

Transcribing Selected Images in a Folder

Open the command palette and run Transcribe images in folder…
Pick a folder from the fuzzy search modal
A checkbox list shows all images found in the folder (all selected by default)
Use Select all / Deselect all to toggle, or check individual images
Click Transcribe to process the selected images

Batch Folder Transcription

Right-click a folder in the file explorer and select Transcribe all images in folder
A progress notice shows the current file being processed (e.g. “Transcribing (3/12): photo.png”)
Up to 3 images are processed concurrently
When complete, a summary notice shows how many succeeded and how many failed
Individual failures do not stop the batch — all images are attempted

Subfolder handling

By default, only images directly in the selected folder are processed. Enable Include subfolders in settings to also process images in nested folders.

Overwrite behavior

By default, images that already have a corresponding .md file are skipped. Enable Overwrite existing files in settings to re-transcribe and update existing output files.

Supported Image Formats

png, jpg, jpeg, gif, bmp, webp, avif, svg