Public Universe
Fair use
This site and its catalogue are free for everyone. Please take what you need for reading, teaching and research, and use an API key when a program does the asking.
Free for everyone
You can browse Public Universe, follow links, and read the pages without an account. Generous anonymous limits cover ordinary visits and light API use. The website itself is not a bulk download path.
Anonymous API limits
Calls with no key are limited per IP address: 60 requests per minute and 1,000 per day. That budget is for people and small scripts. It is not a feed for a crawler that walks the whole catalogue.
Bulk and automated use needs a key
Programs that download many records, mirror the catalogue, or call the API on a schedule must send a key in the X-PU-Api-Key request header. Do not put the key in a query string. A free key has a higher daily limit. An institution key is for universities, schools, museums and similar organisations, and defaults to 100,000 calls per day for that key. That ceiling is a configuration value and may change.
A missing key, a blank header, an unknown key and a revoked key are all treated as anonymous traffic. They use the per-IP anonymous limit, and bad keys are also rate-limited per IP so guesses cannot run unchecked. The API does not answer those calls with a special “invalid key” error. When a limit is exceeded the response is HTTP 429 with a Retry-After header and a link back to the access form.
Cloudflare may skip a stricter anonymous edge rule when a request carries X-PU-Api-Key. That does not make the key valid: the API still checks it. Separately, Cloudflare Pro caps all traffic, including keyed calls, at 300 requests per minute per IP. A key raises the daily allowance. It does not raise that per-IP burst ceiling. An institution that needs a higher burst should list its egress IP ranges on the access form so they can be allowlisted at the edge.
Search and AI crawlers are welcome
Verified search engines and recognised AI crawlers may read the public pages. robots.txt does not block them, and it does not block the sitemap or itself. Please use the sitemap instead of guessing URLs, and prefer the API or the database download when you need the catalogue in bulk rather than the HTML pages.
Interactive MCP clients can keep using the public MCP endpoint without a key. That endpoint is for conversations and small lookups, not for scraping the catalogue.