In Capacities, images are not just extras that you add to your notes. They are their own entities. Open the image object type in the left sidebar to see them all.

All images are image objects, and this means they can individually tagged or enriched with notes.

You can reuse images in any object, with the usual @ or [[]] shortcuts.

Image properties

Every image object in Capacities comes with a set of properties that help you describe, organize, and query your images:

• Title – The name of the image object.
• Tags – Labels you can assign to group related images.
• Description – A text field for notes about what the image contains.
• Blocks – Where you write your notes on the image.
• Source – Indicates where the image came from (for example, upload or Unsplash).
• URL – The source link stored for uploaded image references.
• Category – A predefined classification used for filtering and queries.
• Colors – A palette of dominant colors used for filtering and visual organization.

Believers with AI enabled can let media analysis or other AI features help fill these properties automatically. This an early access feature.

Users on the Capacities Basic plan or Pro plan can still use the properties but it will be a manual action.

Capacities Pro users can use these properties to build queries. Image Analysis will be moved into the Pro tier after the early access period.

Supported file types

The following table shows all supported file extensions for the respective media types:

Adding an Image to Capacities

Simply copy and paste an image into your notes. Or you can type '/image in your notes and choose/upload one from there.

You can also access this option from going to your images object type, finding the three dots, clicking upload and then following the instructions there.

Here, you can drag and drop, paste, browse Unsplash or add an image from a URL.

Different Views for the images

If you are using an image in your notes, you can choose how it looks when embedded into that page.

Image analysis

Image analysis is part of Capacities' media analysis feature. You can use it to analyze your images and automatically extract useful information. Open any image and click the Analyze button (the ✦ icon). The AI will process the image once and fill in the following:

• Title – A descriptive name for the image.
• Short description - A summary of what the image shows
• Extracted text (OCR) – Any text found in the image, including handwritten text. This is rendered as copyable markdown, so text trapped inside images becomes searchable and reusable.
• Category – One of a selection of predefined categories (e.g. document, screenshot, whiteboard, receipt, photo of people, etc.), written to the image's Category property.
• Color palette – Dominant colors shown as visual swatches. Click any color to copy its hex value. These are generalized and then written to the image's Colors property.

All extracted data is written to the image's properties automatically, making your images queryable and filterable across your space.

Image analysis also includes smart title updates: if an image still has a generic title such as “image” or “screenshot”, analysis will suggest a more descriptive title. If you have already given the image a meaningful title, it will not be changed.

Currently, to use image analysis, you need an active Capacities Believer subscription with AI features enabled. Image analysis only runs when triggered by the user.

Offline Image Access

All the images you upload to Capacities will be downloaded to your device and therefore available offline.

If you expect to use a lot of images offline, we recommend uploading a copy of the image rather than referencing it with a link.

FAQ

• How do Category and Topic work with collections?
  If you've been using collections to categorize content, you can now let the built-in Category and Color properties do that work automatically with queries. For example, instead of maintaining a collection for "all images of handwritten notes", you can create a query on Images filtered by the Category property.
  Collections are still useful when you want to manually curate items (for example, "my favorite pics"). But for rules-based groups such as "all images of handwritten notes", queries based on Category/Color can replace those collections and keep your setup cleaner. For more on this, see Can I customize the options for Category, Colors, and Topics?.
• Can I add custom image sections like "blue images" or "architectural pictures"?
  Yes. Beyond the built-in image dashboard sections, you can add your own sections using queries.
  Open the Images object type, click the + next to dashboard sections, and create a new query. Then set filters for what you want to see (for example, Colors includes blue, or Category match architecture). This will be saved as a dashboard section automatically.
  Think of each section as a saved query: you control both what appears in it (filters) and how it appears (list/kanban, grouping, sorting, and hiding empty groups).
• Do I need to use AI image analysis to use the Category and Color properties?
  No. You can fill both properties manually, or use AI image analysis to fill them automatically.
• Why can’t I attach an image directly in AI chat for OCR?
  OCR for images currently runs through image analysis on the image object itself. Direct image attachment in AI chat is not supported right now.
• How can I change the orientation of the image?
  You can switch between landscape and portrait mode by click on the button next to the arrows under the image.
  
• Can I customize the dropdown menus for the Category and Color properties?
  This is not possible and not planned. You can request a new Category to be added on our feedback board.
• How do I change the image title?
  The default title for images is 'image', but you can change that by editing the title when you open your image object. Alternatively, you can use image analysis to rename your images.