zipstream lets you list and download individual files from large .zip archives hosted remotely, without downloading the whole thing. Supports ZIP64, compressed entries, and direct decompression into memory or disk.
Built originally for processing 7TB+ archives like OpenWebVid, it's now available for any data-intensive use case.
- 🧠 Parses ZIP64 central directory via
Rangerequests - 🪄 Supports random access to files in massive archives
- 🔐 No temporary storage required unless you want it
- ⚙️ Works with compressed (
deflate) and uncompressed entries - ☁️ Ideal for streaming datasets, indexing, and lazy data loaders
from zipstream import ZipStream
z = ZipStream("https://example.com/mydata.zip")
file = z.get("videos/000123.mp4")
file.download(base_path="downloads/")