Belluxx/Aleph

Give satellite imagery, terrain data, buildings, streetview photos to your LLM

What it solves

People or LLMs that need information from the physical world (buildings, elevation, photos) usually have to visit many separate services. Aleph lets you draw a shape on a map and download all those layers in one go, saving a full 'piece of the world' to your laptop. It was born to help LLMs gather real-world context quickly, but works just as well for humans.

How it works

You install it with Python 3.11+ and only two dependencies (Pillow and numpy), with no API keys needed. It offers two interfaces:

  1. Dashboard: run alephgeo dashboard, which opens a map in your browser. You draw a rectangle or polygon, pick which layers you want (buildings, elevation, HD satellite imagery, Street View photos, textured 3D models), check the preview and time estimate, and start the capture. Saved captures can be reopened to toggle layers or explore the 3D mesh and terrain.
  2. CLI: run commands like alephgeo streetview --place "Colosseum, Rome" --best-match -o captures to fetch a specific point on the fly.

Who it's for

It's for anyone who wants to keep a copy of a geographic area, or for developers feeding physical-world data into an LLM. The README strongly cautions that Google-sourced data (Street View, satellite, 3D meshes) goes against Google's terms, so you are responsible for how you use what it downloads.

Highlights

  • Draw a rectangle or polygon to capture everything inside it.
  • Bundles multiple data layers: buildings, elevation, HD satellite imagery, Street View photos, and textured 3D models.
  • Works through both a browser dashboard and a command-line interface.
  • Minimal setup — only Pillow and numpy, no API keys, Python 3.11+.

Related

  • Project
    jerryjliu/docjev

    A document classification and splitting tool that uses OCR and natural-language rules to categorize files and divide large PDF packets into separate documents.

  • Project
    conjuncts/gmft

    gmft is a lightweight, high-performance PDF table extraction tool that uses Microsoft's Table Transformer to convert PDF tables into structured formats like Pandas dataframes.

  • Project
    jannisborn/paperscraper

    paperscraper is a Python package that automates retrieval of scientific paper metadata, full texts (PDF/XML), and citation data from sources like PubMed, arXiv, and preprint servers, plus tools for meta-analysis such as self-citation scoring.

  • Project
    supervisely/supervisely

    Supervisely is a web‑based computer‑vision platform that combines data labeling, model training, inference and collaboration tools. It offers a Python SDK and REST API for automation, and an app ecosystem where developers can create head‑less scripts, interactive UI tools, or labeling‑tool extensions—all deployable with a single click.