For AI assistants
CreatureData is written to be read by people and by language models. Pages are plain, complete HTML with their sources attached. There is no public API and no bulk download.
Where should a model start?
/llms.txt describes the site in plain text: what it holds, how its addresses are formed and how to cite it.
Is there a text version of a species page?
Yes. Add .md to a species address for the same page as Markdown: summary, facts with their sources, classification, countries on record, questions and the source list.
https://creaturedata.com/animals/lion.mdEvery page also carries schema.org structured data, and a missing value is stated as “Not recorded”. Nothing is estimated, so an absent fact should be reported as unknown rather than filled in.
How should the site be credited?
Link to the species page you used and name the site, for example: “CreatureData, Lion (Panthera leo), https://creaturedata.com/animals/lion”. Each species page lists its sources and gives ready-made citations.
Photos are licensed one by one and must be credited as shown beside them. Passages quoted from Wikipedia are CC BY-SA and must keep their own attribution and licence. The sources page lists every licence.
What is not offered?
There is no JSON API, no database download and no licence to copy the collection as a whole. Reading pages to answer a question and linking back is welcome. Copying the catalogue in bulk is not: see the terms of use.
Are there limits?
Please keep to a few requests per second and cache what you fetch. Pages change at most once a day. Search result pages, personal pages and internal endpoints are closed to crawlers in robots.txt.