Data endpoints and URL patterns
Reference data is available as JSON, and map state is fully expressible in a URL. There is no server-side generation API — world generation runs in the client.
JSON endpoints
These are static, build-time generated reference documents. They are cacheable and safe to fetch directly.
| Endpoint | Contents |
|---|---|
| /api/versions | Supported Java Edition versions, their release dates and generation profiles. |
| /api/structures | Structure registry: dimension, minimum version, region placement values and required checks. |
| /api/biomes | Biome registry: dimension, category, climate, introducing version and map palette. |
| /llms.txt | Plain-text summary of the site for language models and agents. |
| /sitemap-index.xml | Sitemap index. Sections are served from /sitemaps/pages.xml, /sitemaps/reference.xml, /sitemaps/guides.xml and /sitemaps/seeds.xml. /sitemap.xml redirects here. |
URL patterns
| Pattern | Meaning | Indexed |
|---|---|---|
| /seed/<seed> | Seed map on the latest supported version. Canonical. | Curated seeds only |
| /seed/<seed>/<version> | Seed map pinned to a version. Canonical. | Curated seeds only |
| /seed/<seed>?x=&z=&zoom=&dimension=&layers= | Permalink to a specific view. | No — noindex, canonical points to the clean path |
| /biomes/<biome>, /structures/<structure>, /versions/<version> | Reference pages, one per entity. | Yes |
Permalink parameters
dimension—overworld,netherorend.x,z— camera centre in block coordinates, clamped to the world border.zoom— power-of-two exponent for pixels per block, from -9 to 2.layers— comma-separated structure identifiers. Unknown identifiers are dropped.
Seed values
A seed is a signed 64-bit integer from -9223372036854775808 to 9223372036854775807. Text seeds are converted with Java string hashing before use, so /seed/glacier and /seed/108181935 describe the same world; the canonical form is always numeric.
Usage limits
- World generation happens in the browser, so there is no per-seed rate limit to respect.
- JSON endpoints are static; please cache them rather than polling.
- Automated crawling should respect robots.txt and avoid permalink query URLs, which are noindex duplicates of canonical pages.
Contact
Questions about the data or requests for an additional endpoint: jantrahtech@gmail.com.