# Cesium-Examples: two hundred native CesiumJS demos you can read in one file each > A Chinese-language collection of plain CesiumJS examples, from terrain excavation and viewshed analysis to video projection and particle effects, published as static files with no build step and no wrapper library. **jiawanlong/Cesium-Examples** — CesiumJS2026年最新Examples，200多个原生示例：自定义纹理、地形开挖、可视域分析、洪水淹没、缓冲分析、日照分析、方量分析、通透分析、剖面、单体化、视频融合、视频投射、雨雪雾、天空盒、场景出图、体积云、闪电、视锥体、雷达扫描、水波纹、动态扩散点、电子围栏、粒子效果、物理引擎、克里金、海流、大数据可视化、坐标转换、相机、飞行漫游、瓦片加载、xyz、tms、wms、wmts、天地图、高德、百度、wkt、shp、geojson、pbf、地形、entity、模型、海量数据、Primitive、gif、聚合、czml、字体图标、运动、3Dtiles、模型偏移、等等。。 - Repository: https://github.com/jiawanlong/Cesium-Examples - Website: https://jiawanlong.github.io/Cesium-Examples/ - Stars: 2,211 · Forks: 496 - Language: JavaScript - License: not declared - Published: 2026-10-07 · Updated: 2026-10-07 · Language: en - Canonical page: https://hysenlabs.com/projects/jiawanlong-cesium-examples ## The pitch is that nothing is wrapped The repository description opens by naming the year and the count: CesiumJS examples from 2026, more than two hundred of them, each one native. The README then states the advantage it is selling, and the argument is about what is absent rather than what is present. No plugins, no encapsulation. That is the whole pitch. The comparison the author draws is with SuperMap's i3D and mars3d, the two commercial and open platforms that wrap Cesium with higher-level components, configuration files and a plugin architecture. The claim is that this collection reaches similar visual results while showing you the original Cesium calls, and that every example is one standalone file. Whether you agree with the framing or not, it explains the shape of the repository: no package manifest, no framework layer, no configuration schema, just HTML pages that instantiate a `Viewer` and add things to it. The project has 2,211 stars and 496 forks with only 6 open issues, which tells you something about how people use it. Nobody files many issues against a gallery of demos they are reading for reference. The default branch is `main`, GitHub reports the language as JavaScript, and the last push was 2026-07-28. One licensing point needs stating early. The repository metadata reports no licence, so there is no MIT or Apache grant you can rely on. The README asserts the opposite of what a licence file would, promising the examples are free, carry no charges and need no permission from the author. An assertion in a README is not the same as a grant, so if you plan to put any of this code into a product, ask first. ## What the function list actually covers The README groups the examples into thirteen categories, and the grouping is more informative than the count. Composite applications come first, covering smart city, typhoon tracking, smart traffic, smart logistics, two and three dimensional linkage, aerial survey, and running Cesium together with Three. Basic features follow: coordinate work and coordinate transforms, camera control, measurement, events, fly-through roaming, compass, navigation bar and timeline. Imagery layers show how to add a single image, an xyz service, tms, wms, wmts, Tianditu, AMap and Baidu. Vector layers cover shp, geojson, mvt, czml and animated time-series layers. Terrain gets its own entry, with offline terrain, online terrain and terrain exaggeration. The Entity examples cover points, polylines, polygons, models, massive icon sets, popups, gifs, clustering, czml, font icons and entity manipulation. Primitives are collected separately, including PrimitiveCollection. Model examples handle glb, gltf and 3D Tiles loading, model events, model editing, model movement, model offsetting, custom materials and custom shaders. Spatial analysis is the category that most rewards the no-wrapper approach, because these are the operations that are painful to write from scratch: terrain excavation, viewshed analysis, inundation analysis, buffer analysis, sunlight analysis, volume calculation, perspective analysis, profile analysis and single-building extraction. Video fusion follows, covering video overlay, video projection, two and three dimensional projection and ground-hugging projection. Effects take the last big block, with rain, snow and fog, sky boxes, scene imagery export, volumetric clouds, lightning, frustum cones, radar sweeps, flow lines, water ripples, dynamic diffusion points and geofencing. Particle effects include flame, water jets, explosion, spray and smoke. A physics section covers vehicle motion, rotating objects such as flags, soft-body collisions and fracture. The final category is visualisation: large-data rendering, wind fields, 3D and 2D heat maps, kriging, ocean currents and isosurfaces. ## Running them means hosting static files, nothing more The entire deployment instruction is one sentence: once you download the project, put the complete folder under a container such as nginx. There is no build, no package install and no bundler step. The repository tree confirms that shape, with an `index.html` for the landing page, a `code.html` and `code.js` pair that the index presumably uses for the browser-side source viewer, an `examples/` directory holding the individual demos, a `libs/` directory for vendored dependencies, and a `donate/` directory. That is genuinely the whole setup, and it cuts both ways. The upside is that you can open one file, read it, change one call, and see the result. There is no abstraction to trace through first. The downside is that you are responsible for everything the wrapper would have handled: matching the Cesium build in `libs/` to the version your tile provider expects, supplying an access token where required, pointing terrain and imagery at services you have credentials for, and choosing the right origin settings for the data sources you load. The README also links a second repository, a low-code Cesium platform, and a separate collection of more than a hundred Three.js examples, plus a site described as Cesium Skills. Those are adjacent projects by the same author rather than parts of this one, but they suggest the same approach applied at different levels: show the raw API first, then offer a higher-level tool separately if you want it. A live demo site is published on GitHub Pages at the repository's own URL, which is the fastest way to see whether the examples still run against current tile providers. That matters, because a static gallery of browser examples silently rots when an imagery endpoint changes its terms. ## Screenshots, community links and a long donation table The README ends with two screenshots, contact details for the author through WeChat and a QQ group, and then a donation table that runs from February 2025 to July 2026 with around eighty entries. It is worth describing that section rather than skipping it, because it tells you how the project is sustained. There is no release cadence, no published versions and no sponsorship platform. What there is a claim of near-continuous updating, with new cases arriving most days or weeks, and a note that donations are forwarded on in the donor's name to a charity, first to Tencent's charitable giving and then, from January 2026, to a named student. A spreadsheet of the details and a certificate page are linked from the repository. The author is explicit that this is how the work gets funded, and the donate directory in the tree is where that lives. If you are weighing whether to depend on this repository, the practical question is not whether the examples are any good, since the function list suggests they are substantial, but whether an unpinned, unversioned, unlicensed collection is something you want as a dependency. The safer pattern is to read it, extract the two or three calls you need, and write them into your own codebase. ## Where this fits against official docs and commercial platforms Cesium itself has a sandbox and a long feature list, and any honest comparison has to start there. What the official examples tend to do is demonstrate one API surface at a time in a minimal context, which is right for learning a single call and less useful for seeing how a finished effect is assembled. A project like this inverts the ratio: the individual call is easy to find, and the assembly is what you are copying. The commercial platforms occupy the other end. They give you a consistent project structure, layer management, entity binding, a plugin ecosystem and a UI layer for picking data sources, and in exchange they ask you to learn their abstraction and accept their version coupling. That is the right trade when you are shipping an application with a team and a maintenance budget. It is the wrong trade when you want to know how volumetric clouds or a geofence actually work. The most valuable use here is as a reference implementation with a specific question in mind. You have a spatial analysis or video fusion effect to build, you search the category list, you open one file, and you read the Cesium calls that produce it. The rest of the project, including the absence of a licence, matters only if you are planning to ship rather than learn. There is a geographic caveat worth stating as well. A meaningful share of the imagery examples target Chinese map services, and the visible examples are documented in Chinese. Neither is an obstacle to reading the code, but if you need a vendor-neutral setup or documentation in English, the fit is partial and you will spend more time translating intent than you would with an English-first gallery. ## Conclusion The reason this repository is useful is structural rather than technical. Cesium's documentation explains the API and commercial products such as mars3d and SuperMap wrap it in a larger framework, so when you are stuck on a specific effect you often end up reading two layers of abstraction to find one line of Cesium code. Here each example is a single file with the raw API visible, which makes it a lookup table rather than a tutorial series. The trade-offs are equally plain. There is no declared licence on the repository, so the default copyright applies and you should confirm terms before shipping anything derived from it. There are no releases, so nothing is versioned for you. The scope is Chinese-language tooling and Chinese map services, which narrows who finds it immediately useful. And every example assumes you can host static files and supply your own terrain, imagery and token. If you can do that, this is the fastest way to see what Cesium can actually do before you commit to a project structure. ## FAQ ### What licence does Cesium-Examples use? The repository metadata reports no licence, which means the default copyright applies rather than an open source grant. The README states that the examples are free to use with no charges and no need to ask the author for permission, but that is a claim in documentation rather than a licence file. Confirm terms with the author before shipping code derived from it. ### How do I run these examples locally? Serve the downloaded folder from any static file server. The README's whole deployment note is to place the complete project under a container such as nginx. There is no build step, package manifest or bundler, because each example is a standalone file. You are responsible for supplying terrain, imagery services and any access token your data sources require. ### Do the examples use plugins or a wrapper library? No. The repository's stated advantage is that every example is native CesiumJS with no plugins and no encapsulation, and the README positions this as its difference from platforms such as mars3d and SuperMap's i3D. That also means there is no higher-level layer to configure, and every Cesium call is visible in the file you open. ### Does the repository have versioned releases? No GitHub releases are published for this project, so there is nothing to install through the releases page and no stable version to pin. The README describes ongoing updates, with new cases arriving most days or weeks, and the canonical reference is the live demo site plus the repository contents on the main branch. ### Which spatial analysis operations are demonstrated? The README lists terrain excavation, viewshed analysis, inundation analysis, buffer analysis, sunlight analysis, volume calculation, perspective analysis, profile analysis and single-building extraction. Alongside them are video fusion and projection examples, Entity and Primitive collections, model work including custom shaders, weather and volumetric effects, particle systems, a physics section, and visualisation examples for wind fields, heat maps, kriging and ocean currents. ## Sources - [Issues](https://github.com/jiawanlong/Cesium-Examples/issues) - [jiawanlong/Cesium-Examples on GitHub](https://github.com/jiawanlong/Cesium-Examples) - [Project website](https://jiawanlong.github.io/Cesium-Examples/) - [README](https://github.com/jiawanlong/Cesium-Examples/blob/main/README.md) --- Hysen Labs editorial analysis, written from the project's own repository and release notes. Cite the canonical page: https://hysenlabs.com/projects/jiawanlong-cesium-examples