God's Eye View is a browser-based spatial-intelligence app that layers live public data feeds onto a photorealistic 3D globe, so a laptop behaves like a satellite operator's console. It is built for people who want to watch the world in real time without a clearance or a subscription: open-source intelligence hobbyists, reporters and analysts who follow flights and vessels, geospatial developers looking for a full working globe application to read and extend, and anyone who just wants to spin the Earth and see what is moving on it right now. The project comes from the creator behind the God's Eye View video series, formerly called WorldView, and the whole thing is open source.
What it does
The globe is the interface. On top of the terrain the app draws the signals that public sources already publish, and the README names them directly: aircraft, ships, satellites, earthquakes, traffic and public cameras. Nothing here is a rendered simulation of those objects; the pitch is that the imagery looks like a spy tool while the feeds underneath it are the same ones anyone can read.
From there it leans into the fantasy on purpose:
- Click any aircraft and ride inside the cockpit, following the flight from the pilot's seat as it moves.
- Switch the view to thermal, night vision, or a military-style heads-up display over the terrain.
- Talk to it. Hands-free voice control is handled by a realtime AI agent, and in the demo the spoken instruction is enough to make the app draw boundaries on the map for you.
That combination — real feeds, cinematic presentation, voice as a first-class control — is what separates it from the usual flight-tracking or ship-tracking web page, where you get a flat map and a table of callsigns.
How it works
It is a JavaScript project that runs in the browser on WebGL, and the repository's own topics point at the stack: Cesium for the 3D globe, photogrammetry for the terrain, and a spread of geospatial and tracking concerns around them. The live layers are pulled from open sources rather than from a proprietary backend, which is the reason the project can be handed out as a clone-and-run repository at all — there is no service behind it that someone has to pay for on your behalf.
Voice is the one piece that reaches outside the browser, since it is driven by a realtime AI agent. Optional API keys unlock parts of the experience, and the README is explicit that you add them inside the app after it is already running, with a dedicated section covering keys and their costs.
Getting started
The headline promise is that you start without API keys. Two paths are offered: install it through Pinokio, which the README links to directly, or clone the repository and run it locally from the terminal. The video makes the same point — clone it and run, no accounts required — and optional keys go in later through the app's own settings rather than a config file you have to prepare in advance.
The README is organised for that first session, with a quick start, a "first five minutes" walkthrough, a guide to talking to the app, an inventory of what is live on the globe, and an under-the-hood section for people who want to understand the rendering and data plumbing before they modify it.
When to use it / when not
Reach for it when you want situational awareness that is visual rather than tabular: watching air traffic over a region, checking recent earthquakes in context with terrain, following vessel movement, or showing someone what open-source intelligence actually looks like in practice. It is also a strong reference implementation if you are building anything on Cesium and want to see a large, polished globe application with live data layers and an agent-driven control surface.
It is the wrong tool when you need an authoritative or complete picture. Everything on this globe comes from public feeds, so anything that does not broadcast publicly is simply not there, and gaps in coverage are gaps in the underlying sources, not bugs in the app. It also wants a live connection and a machine that can render a photorealistic globe comfortably, so it is not an offline or low-power option. And if you want the voice control and every optional layer, budget for the keys the README discusses rather than assuming the free path covers all of it.
Take this one seriously if you work anywhere near geospatial data, open-source intelligence, or 3D on the web. The reception backs that up: the repository reached number one on GitHub Trending, daily and weekly, in August 2026, and has gathered over 22,000 stars and 4,700 forks in the months since it was created — unusual traction for a project this young. For everyone else it is still worth an evening, because it is the rare piece of open-source software that makes a genuinely impressive first impression and then turns out to be honest about where its data comes from.