Many catalogs.
One search.
Find satellite imagery across providers with one query. Combine the results, save the metadata you need, and explore it again with fewer API requests.
pip install superstacsearch()Find imagery. Keep the metadata you need.
One search workflow for Python notebooks, Rust applications, and the command line.
Search across providers
Query relevant catalogs concurrently and receive combined results, with visibility into which sources responded or failed.
How federation worksSave once. Explore repeatedly.
Save your study-area metadata to GeoParquet and reuse it across searches. Automatic mode checks live catalogs when the saved data is too old or doesn’t cover your query.
Build your inventorySpend less time combining results
Configure collection and asset aliases, remove duplicate item IDs, and pass STAC metadata into your existing analysis workflow.
Run your first searchKeep exploring the same study area.
Exploring the same region over multiple sessions? Save its catalog metadata to GeoParquet, then refine your searches locally. Scope ingestion by collection, area, and time so you only collect what you need.
- Follow ingestion progress, set a storage limit, and resume interrupted work.
- Search the saved coverage locally; let automatic mode fall back to live catalogs when needed.
- Add recent dates and clean up old files as your project grows.
Inventories store metadata. Imagery previews and downloads still use the provider’s assets.
Try the GeoParquet notebookExplore different queries within your saved coverage.
How fast is local search?
Our benchmark measures how skipping irrelevant data speeds up a selective search over 100,000 synthetic items. See the results and run it yourself.
Explore the benchmarkGuides
Search example
Register the catalogs you want to search, then query a collection. SuperSTAC sends the request to the relevant catalogs and combines their results.
- Filter by location and date.
- Normalize collection and asset names, and remove duplicate item IDs.
- Check which catalogs responded and which failed.
from superstac import Client
client = Client(catalogs=[
{"id": "earth-search", "url": "https://earth-search.aws.element84.com/v1"},
{"id": "microsoft", "url": "https://planetarycomputer.microsoft.com/api/stac/v1"},
])
try:
result = client.search(collections=["sentinel-2-l2a"], limit=5)
for item in result.items():
print(item["id"])
finally:
client.shutdown()Why SuperSTAC?
The pipeline problem that led to SuperSTAC, and the thinking behind it.
SuperSTAC is in alpha. APIs and configuration may change. Current limitations →