Seeder
(1.5.0)Who
I'd love to know if you use this website and how
Can I help?
What is it
Find, show, share and admire Minecraft seeds (Java edition mostly, Bedrock only if you are interested in biomes).
Works with every Java version from Beta 1.7 to 26.3.
This is a webapp based upon a library (cubiomes) written by cubitect in C and compiled in webassembly by me. Go and give him love. Since the original library is no longer maintained, Seeder now builds against the actively maintained xpple/cubiomes fork.
It uses fancy new technology to run the same code in your browser, so that you can look up your seeds/find your perfect ones without installing or configuring anything.
Found a bug?
Report bugs and ideas on GitHub.
What it does
- Draws the biome map of any seed, in the Overworld, the Nether and the End (drag to pan, mouse wheel or pinch to zoom)
- Shows biomes at any height from 1.18
- Finds the spawn point, the strongholds and the structures
- Slime-chunk and chunk-grid overlays on the map
- A dashboard next to the map for every seed:
- spawn, and strongholds with the number of eyes in their portal room
- the nearest structures, with their variants (zombie villages, giant ruined portals, bastion types, End city ships…)
- which biomes are around spawn, and where the nearest patch of any biome is
- Find near me: what is at any coordinates you type
- farms: whether the seed has a quad witch farm, and blaze spawners in fortresses
- My worlds: save the seeds you play
- Finds seeds with several biomes and several structures at once, many results per search
- Presets for common searches, led by what the newest Minecraft version added: survival starts, speedrun, stacked structures, builders' biomes, structure hunts, the Nether and the End
- Share a seed or a search with a link, export results as CSV or JSON, or share a picture of a seed
- Finding rare seeds may slow down your PC while the search is running (seed finding is a very CPU intensive job)
Limitations
- Seeder reproduces Java Edition. From 1.18, a seed gives the same terrain and biomes on Bedrock Edition, so the biome map works for Bedrock worlds too; structures, strongholds and slime chunks are placed differently on Bedrock, so those markers only hold for Java.
- From 1.18, desert pyramids, jungle temples and mansions need high enough ground, which cubiomes can only estimate: a few of the ones shown may not generate.
- The map's terrain shading (from 1.18, in the Overworld) is an estimate of the ground: mountains read lower than in game (about YÂ 170 at most), and cliffs and overhangs do not show.
- The number of eyes in a stronghold's portal room is unknown before 1.13.
- Dungeon and mineshaft spawners depend on terrain carving and cannot be predicted from the seed.
- Every finder search is centred on the world origin (XÂ 0, ZÂ 0), not on the spawn.
- With a biome in it, a finder search reaches at most 1,000 blocks from the origin.
- Biome searches are slow from 1.18: within 300 blocks, one core checks about a dozen seeds per second.
Credits
- World generation: cubiomes by Cubitect (MIT licence), via the xpple/cubiomes fork.
- The pixel font is Minecraftia by Andrew Tyler, licensed CC BY-SA 3.0.
- The circle generator's circles follow donatj's Circle Generator (MIT licence), and its spheres are inspired by WorldEdit's //hsphere.
- The full licence texts are in licenses.txt.
Privacy and ads
Seeder has no accounts and no server of its own: maps and searches run in your browser. Google AdSense shows ads, and Google Analytics tells me how the site is used, so I know what to fix and what to build next. Like any page view, they see the address of the page you are on, which includes the seed, version or search criteria.
What Analytics collects, and why
| What | Why |
|---|---|
| The pages you open, with their address (seed, version, search criteria) | To know which pages, versions and dimensions people use |
| Which features you use: dashboard tabs, copy and share buttons, map tools, the map's zoom, Go to and Return to spawn buttons (never the coordinates), saving a world (never its name) | To see what is worth improving, and what nobody uses |
| Which tools you use and which of their buttons you press (never what you type in them) | To know which tools people actually use, and which ones to build next |
| Finder searches: the preset, or the structures, biomes and spawn island size asked for and the world type, and how the search ended (seeds found, seconds) | To tune the finder and its presets |
| Speed: how long the map engine takes to load, a range of CPU cores, page speed and which part of the page moved while loading (an ad, the map, the panel), engine errors | To catch slow or broken devices |
| Whether an ad space stayed empty and showed the support note instead | To know how often ads are blocked or unfilled |
| Clicks on the support links in an empty ad space (PayPal or Buy me a coffee), never amounts | To know whether the note is worth its space |
| Country, device type and browser, as Google Analytics records them for every site | The standard reports |
Never collected: world names, coordinates you type, anything else you write, or who you are. Google Analytics 4 does not store IP addresses, and the data is kept for 14 months.
The switch covers analytics only. Ads and their cookies follow Google's own consent message where the law asks for it, and "Privacy and cookie settings" at the bottom of every page reopens it to change your choice. Ad personalisation can also be managed in Google's ad settings.