Your first map.
On your machine.
Tellurion v0.5.0 is available under AGPL-3.0-only. Download, extract and start a synthetic GeoPackage demo with the embedded map interface. No Docker, external database or account required.
Download. Verify. Open the map.
- Verify your archive against the release's SHA256SUMS and GitHub attestation instructions.
- Extract it, keep both executables together, and open a terminal in that directory.
- Run
./tellurion-ingest demo --path ./demo.gpkg --port 8080on Linux. In Windows PowerShell run.\tellurion-ingest.exe demo --path .\demo.gpkg --port 8080. - Open
http://localhost:8080/ui/in your browser. The command creates synthetic sample features in a local GeoPackage.
For the raw response, open http://localhost:8080/public/features/catalogs/default/collections/demo/items?limit=10. Stop the server with Ctrl+C when finished.
Then try your data
Use the versioned guides to choose a supported driver and configure your source. Reads, writes and optional build features vary by format; the native package does not include every optional driver.
The separate public-file preview accepts supported public HTTPS sources such as COG, GeoParquet and ZIP Shapefile. It evolves independently and is not a version-pinned v0.5.0 hosted demo. Do not submit credentials or sensitive data.
Administration requires your own identity provider, durable control storage and permissions. This local demo grants no admin access. Read the tenant and catalog administration guide for the scoped cache-lifetime editor and its limits.
A release you can inspect
The published archives were built from commit 000f40adc5b35cec5a4bebdbf3e8d18c64fc01b8. Linux and Windows passed extracted-package startup, feature and UI checks. Build evidence · Source, dependency notices and release notes.
The historical online demos retain their existing versions and links. This page does not relabel those backends as v0.5.0. No Tellurion Cloud, SLA or support service is offered.
Tell me what worked, which format you tried, and what got in your way: report reproducible feedback or contact Carlo Cancellieri on LinkedIn. Keep private datasets and credentials out of reports.