Skip to content


Switch branches/tags

Name already in use

A tag already exists with the provided branch name. Many Git commands accept both tag and branch names, so creating this branch may cause unexpected behavior. Are you sure you want to create this branch?

Latest commit


Git stats


Failed to load latest commit information.
Latest commit message
Commit time

A collection of metadata associated with the collection of the Minneapolis Institute of Art.

Mia's mission is to enrich the community

by collecting, preserving, and making accessible outstanding works of art from the world’s diverse cultures.

To increase access to Mia's outstanding art, we've published metadata for our artworks as JSON under a CC0 license.


This is the same data that we publish on our collections website. Having everything on its own page is a great way to browse, but taken as a whole it's much easier to see the overall shape of the data. For example, if someone wants to know how many artists have work at Mia,

$ find objects -name "*.json" | xargs cat \
	| jq -r '.artist' \
	| sort | uniq -ci | sort -nr

is a simple program that knows. How many of those artists have the same name as you? Save the output of the above command to a file called artsmia-artists.txt, then cat artsmia-artists.txt | grep -i <your name here> | wc -l will tell.

Or maybe you want to use markov chains to make fake descriptions of objects in official museum-speak…

$ find objects -name "*.json" | xargs cat \
	| jq -r '.description, .title, .text' \
	| grep -v '^$' \
	> descriptions.txt
$ dadadodo -c 1 descriptions.txt
Black lines of concentric diamonds (beard architectural African symbols
in the foot; body and black plum blossom petals).

(For these two examples, you'll need jq and dadadodo, along with a sane command line environment.)

Having this raw data in an accessible format opens the doors to experimentation that isn't possible with our collections website alone.



Mia identifies objects curatorially by accession number. Computertorially, we use numeric object ids beacuse they're easier to deal with.

Each record lives at objects/$bucket/$id.json, where 'bucket' is object id / 1000. So 0/ holds records 0 through 999; 1/ holds 1000-1999; etc. Here's what objects/0/17.json looks like:

  "accession_number": "13.59",
  "artist": "Walter Shirlaw",
  "continent": "North America",
  "country": "United States",
  "creditline": "Gift of Mrs. Florence M. Shirlaw",
  "culture": null,
  "dated": "19th century",
  "description": "",
  "dimension": "3 1/4 x 7 1/2 in. (8.26 x 19.05 cm)",
  "id": "",
  "image": "valid",
  "image_copyright": "",
  "image_height": 2263,
  "image_width": 3593,
  "life_date": "American, 1838 - 1909",
  "marks": "Signature [Cheyenne. W. Shirlaw]",
  "medium": "Graphite",
  "nationality": "American",
  "provenance": "",
  "restricted": 0,
  "role": "Artist",
  "room": "Not on View",
  "style": "19th century",
  "text": "",
  "title": "Sketch made on Indian Reservation, Montana"


Images aren't included under the same license as this metadata. Reference images are available under Mia's Image Access & Use policy.

There aren't images of every object in the collection. Of the objects that have been photographed, some are restricted by copyright.

  • "image": "valid|invalid": is there an image available for this object?
  • "restricted": "0|1": is it usable? 0 says that an image is not "subject to any additional terms or restrictions", such as copyright or a special request from the artist to limit display of an object. 1 means the opposite.

Feel free to use unrestricted images for "limited non-commercial and educational purposes". When the copyright owner of an artwork is known, it will be in "image_copyright": "…". Commercial licensing is handled through Bridgeman Images.

Image thumbnails are accessible by their object id in three sizes:$id/{small,medium,large}.jpg. Small images are 100px on the long side, medium images 600px and large images 800px.


Are organized in "buckets" just like objects.

  "exhibition_id": 734,
  "exhibition_department": "Decorative Arts, Textiles & Sculpture",
  "exhibition_title": "Warren MacKenzie: The Legacy of an American Potter",
  "begin": 2007,
  "end": 2007,
  "display_date": "Saturday, May 19, 2007 - Sunday, August 26, 2007",
  "objects": [



Our records are constantly updated and added to. Changes are committed to this dataset approximately once per day.

CC0 (+BY)

The Creative Commons Zero (CC0) Public Domain Dedication allows this data to be reused free of restrictions. (+BY) suggests "implied or ethical attribution".

Please use this data. When you do,


We're happy to hear your thoughts on our data: what you think of it, how you're using it, and if there's anything you'd like to see changed. The best way to communicate in regards to this data is by using github "issues" and "pull requests".

However, this repo is largely generatedº from our upstream "collection management software" (TMS). Making enhancements to the generated json files is a fantastic way to point out better ways we can represent data, or alert us to possible improvements. Such pull requests won't be merged directly, but instead factored into future improvements to our data pipeline.

See also


No releases published


No packages published