Cosmic History Timeline is a free, interactive visualization that places the entire history of the universe — from the moment of the Big Bang, through galaxy formation, the rise of life on Earth, the great civilizations, religions, empires, wars, and modern nations — on a single, zoomable timeline you can explore in your browser. No signup. No tracking. No paywall.
We believe scale matters. A textbook can show you the Roman Empire in chapter four and the cosmic microwave background in chapter twelve, but it cannot show them at the same time — at proportional distances, on the same number line, where 700 years of Pax Romana sits as a thin sliver next to 380,000 years of cosmic recombination. That comparison, that proportion, is what this site exists to make possible.
Open the main canvas and you are looking at 13.8 billion years compressed into a horizontal axis. The leftmost edge is the Big Bang, marked at year −13,800,000,000. The rightmost edge is today, with a soft "future zone" extending five hundred years past the present so you can see history unfinished, still in motion.
Stacked vertically along this axis are more than 250 timeline tracks, each one carrying the events of a single thread: a country, a civilization, a religion, a war, a scientific discipline, a geological eon. The tracks are organized into six groups — natural systems at the top (cosmos, galaxies, the solar system, Earth, life), then global themes, regions, civilizations, religions, and individual nations at the bottom. Colors follow each group's character: cool blues and purples for the cosmic and natural; warm reds and oranges for wars and certain region clusters; deep crimsons and umbers for ancient empires; the green family for world religions; flag-derived hues for modern countries.
On every track you find events, drawn either as single points or as horizontal pills covering a span. A point is a moment — the founding of a city, a battle, a coronation. A pill is a span — a dynasty, a war, an ice age. Click any event and a panel opens with a curated description in both Chinese and English, an image (almost always with full attribution to its Wikimedia Commons source), a Wikipedia link in each language for further reading, and where appropriate, a recommended book. Some events also draw thin curved lines from one track to another: these are cross-references, the visualization saying "this event matters here too." A meteor strike that ends the dinosaurs draws a curve from the geology track to the life track. The Mongol invasions draw curves from a central conquest band into Russia, Persia, China, and Eastern Europe simultaneously.
Zoom out to the maximum and you see cosmic time at its full scale: the universe expanding, galaxies forming, the solar system condensing, life crawling onto land. Zoom in and the canvas resolves into smaller and smaller intervals — first centuries, then decades, then individual years and months. Where the same year holds many events, they cluster automatically; click the cluster and a small menu lets you choose which one to read. The level system controls density: at low zoom you see only the most important (level 1) events; zoom past 0.55 and the next tier (level 2) appears; zoom past 0.95 and every detail is visible. This way the canvas never overwhelms — it reveals progressively as you focus in.
This is where a single zoomable canvas does what a textbook cannot.
Set the time window to 500 BCE. On the Greek track you find the Peloponnesian War brewing and the early career of Socrates; directly above it on the Persian track, the Achaemenid Empire is at its fullest extent; on the Chinese track, the Warring States period is producing Confucius's first generation of disciples; on the Indian track, the Buddha is teaching; on the Hebrew track, the Second Temple has just been rebuilt. Five world traditions, all alive at once, each shaped by what was happening on the others' tracks. The canvas places them at the same vertical line so the simultaneity is visual, not paragraph-by-paragraph.
Pull the zoom out to see human civilization — every empire, every war, every dynasty — collapse into the rightmost one-twentieth of one percent of the total timeline. Most of the canvas, by area, is the early universe and Earth's deep geological eras. The visualization makes the asymmetry impossible to ignore.
When the Western Roman Empire falls (track: Roman Empire), trace the cross-reference curve and you'll see it lights up the Byzantine track, the Christianity track, and the Germanic-tribes track — three institutions whose centuries-long divergence begins from a single moment of dissolution. Black Death events cross-reference into trade, into demographic collapse, into religious upheaval. The canvas treats history as networked, not siloed.
Through the country views you can pull up Japan and Korea, or Britain and France, or Egypt and Iran as parallel tracks. Their conflicts, treaties, dynastic intermarriages, and parallel revolutions become a visual pattern instead of two unrelated reading lists.
Display the Five Ice Ages view alongside the 20th-century wars. The Cryogenian glaciation pill spans tens of millions of years; World War II spans six. They share the same horizontal axis. The canvas refuses to flatter human time.
As of late April 2026, the dataset is roughly the size of a small encyclopedia and growing weekly. Every number below reflects what is currently in the live events.json; the figures rise as new events are added.
Each of the 2,700+ events carries: a year (or a year range, for spans); a primary track; a level of importance from one (canonical) to three (detailed); titles and 50–200-word descriptions in both Chinese and English; usually an image with full author and license attribution; usually a Wikipedia link in each language; and frequently a recommended book with full title and ASIN. About 99.6% of images include explicit author and license metadata — fetched from Wikimedia Commons via its public API and stored alongside the event so the credit shows on every page where the image appears.
The dataset is curated, not exhaustive. We do not aim to list every event; we aim to choose the events that, taken together, give a faithful proportional picture of each thread. Each new addition is reviewed for source quality, period balance, and visual coherence on the canvas.
Events are organized into eight thematic sections, accessible from the main menu. Each section contains multiple "views" — preset combinations of tracks, year ranges, and zoom settings — designed for a particular kind of comparison or narrative. Numbers in parentheses are current event counts.
The biographies section is the next major area under development: each figure will receive a curated personal timeline plus references back into the political, military, and cultural tracks they shaped.
In addition to the interactive canvas, every view with substantial content has a corresponding reading page — a long-form, scrollable, magazine-style version of the same data. The reading page lists every event in chronological order, with image, full description, related axes, Wikipedia link, and a "📍 locate on main timeline" link that will jump you back to the canvas at the exact moment of that event.
Each reading page also embeds the live interactive canvas at the top, scoped to that view's tracks. So you can scroll-read the events and, if a particular era catches your attention, scroll up and explore that era visually without leaving the page.
Currently 79 reading pages are generated, covering all eight sections except the in-progress biographies. The full list lives at views.html — the Reading Index, also accessible from the main canvas via the prominent banner at the top of the section menu.
Every event description is written in both Traditional Chinese and English. The two versions are not machine translations of each other; they are written separately, and each tries to use idiomatic phrasing in its own language. Place names, person names, and historical terms follow each language's standard transliteration (e.g., 凱撒 / Caesar, 鄂圖曼 / Ottoman).
The interactive canvas has a single 中 / EN toggle in the header. Clicking it swaps the entire UI and every event description between languages. The reading pages have the same toggle in their top-right corner. The choice is remembered across pages and visits via your browser's local storage, and is also reflected in the URL (?lang=zh or ?lang=en) so you can share a link in the language the recipient prefers.
When you switch languages, the <html lang> attribute updates accordingly, telling search engines and accessibility tools which language they're reading. Reading pages additionally publish hreflang annotations in the sitemap, so search engines surface the right-language version to the right audience.
Beyond pan-and-zoom, the main canvas has a generous set of features for navigating the dataset. None of them require an account; everything works in your browser.
| Feature | What it does |
|---|---|
| Search | Type any name, place, or year. Instant dropdown of matches; clicking jumps the canvas to that event with the popup open. Searches both Chinese and English titles. |
| Filter by category | Toggle categories on or off (war, science, religion, culture, politics, etc.) to focus on one kind of event. |
| Era jumps | Twelve era buttons across the bottom — Big Bang, Earth & Life, Dinosaurs, Human Origins, Prehistory, Antiquity, Classical, Medieval, Renaissance, Industrial, Modern — instantly zoom to that period. |
| View selector | The "Events" button in the header opens a categorized list of all 88 views. The Reading Index banner at the top opens the long-form text version of every topic. |
| Solo & toggle on legend | Single-click any axis label in the right-hand legend to hide/show that track; double-click to enter solo mode (only that track stays at full opacity, others dim). Click an axis-group title to toggle the whole group. |
| Click-to-fade on canvas labels | Click an axis label on the canvas itself (the sticky labels on the left edge) to fade just that track — useful when something below the label is hidden by it. |
| Pill expand | Some pills (like the geological eons) contain nested child pills — click the parent to expand them. |
| Cross-reference popups | The hollow circles drawn on remote tracks indicate that an event from elsewhere references this track at this moment. Click the circle and the originating event's popup opens. |
| URL deep-linking | Every view has a permalink: /?v=ww2, /?v=japan, etc. Add &year=1945 to land at a specific year. Add &lang=en to force a language. Bookmarkable, shareable. |
| Mobile & touch | Pinch to zoom; one-finger drag to pan; tap for popups. The canvas reflows for smaller screens; reading pages are mobile-first. |
| Auto-update | If you leave the tab open and come back later, the page silently checks for new data and reloads itself if there's been an update. You'll always be looking at the latest events. |
| Era band overlay | Soft horizontal color bands behind the tracks indicate named historical eras (Renaissance, Enlightenment, Industrial Age) so you have a visual at-a-glance sense of "where in history" any vertical slice falls. |
| Future zone | The 500-year band past the present is rendered with a gentle hatch, marking it as the "not yet written" portion of the timeline. Some events are projected here (e.g., known asteroid flybys, scheduled space missions). |
The site is one author's curatorial project: every event is hand-selected, every description is hand-written. There is no committee and no automated content generation; the writing voice is consistent because one person did all of it.
Selection follows three rules: (1) if a topic appears in any standard secondary-school or university-level world-history textbook, it should appear here; (2) events that link two or more tracks (cross-civilization influence, cause-and-effect across continents) take priority over events that stay within a single thread; (3) within a single track, the rhythm should feel proportional — no decade overflowing while the next century is empty. The result is a deliberate, opinionated dataset rather than an exhaustive one.
Description writing follows two more rules: (4) each description should answer "who, where, what, and why does it matter," in roughly that order, in 50–200 words; (5) the phrasing should be writeable as a paragraph, not as a bullet list — readers should be able to actually read the event, not skim a bullet.
Where the curatorial decisions become harder — for instance, which events from a country's full canvas history to include on its dedicated reading page when the canvas has 250+ candidates — we use a large-language-model assist to score topical relevance. The LLM does not write any of the prose. It does not select novel events. It only makes inclusion judgments on a candidate set the curator has already approved, by reading the event's pre-written description and answering "would a reader expecting the topic of this page expect to see this event?" The judgments are stored in the dataset (relevant_views) and reviewed; clear errors are corrected with manual overrides. This kept the reading pages focused without restricting the canvas itself.
The factual scaffolding for events comes primarily from Wikipedia in both Chinese and English. Each event has a Wikipedia link in each language, surfaced in its popup and on its reading-page entry. Where Wikipedia disagrees with itself across language editions (which it does, for many topics), we cross-check against published historical surveys before deciding the version we tell.
Images come almost entirely from Wikimedia Commons, the free-media repository sister-site to Wikipedia. Every Commons image we display has been resolved through the Commons API to obtain its author of record, the license under which it is shared (Public Domain, CC0, CC BY, CC BY-SA, etc.), and a direct link to the file's Commons page. This metadata is shown beneath every image on every reading page, in the format:
A small minority of images come from NASA, which produces them as part of its public mission and releases them into the public domain. NASA images are individually credited where used.
We do not host or rehost images on our own infrastructure. Every image is hot-linked to its original Commons (or NASA) URL. This means the image you see is the same image you would see by visiting the source page, and that any later edits, corrections, or takedowns on the source side propagate to us automatically.
Recommended-reading book references use Amazon's product pages. The book and its ASIN are stored alongside the event description so we can show the actual title an author. We do not display book covers or scrape any content from Amazon; the link simply leads to the Amazon product page where the reader can decide whether to buy. See the affiliate disclosure below for what happens when you click through.
Event prose itself — the 50- to 200-word descriptions in Chinese and English — is original editorial work, written by Cosmic History Timeline. It is informed by Wikipedia and other published sources, but it is not a translation, copy, or close paraphrase of any one of them. Reuse of our event descriptions is governed by the copyright section below.
The site is built as a single-page application with vanilla JavaScript and the HTML5 Canvas API. No frontend framework. No bundler. No backend server. No database. The entire dataset is one JSON file (events.json) loaded once at page open; everything subsequently happens in your browser. This deliberate simplicity has consequences:
We use no analytics that follow you across sites, no behavioral tracking pixels, no fingerprinting. The only third-party JavaScript on the page is Google AdSense — a single ad slot at the bottom of the canvas, and one or two slots on each reading page. AdSense sets its own cookies, which is disclosed in the privacy policy and discussed in the next section.
We do not collect personal information. We do not require an account. We do not use server-side analytics; there is no server-side at all.
Your language preference (Chinese or English) and any local edits you make using the optional editor (a separate page) are stored in your browser's localStorage. They never leave your device.
This site displays Google AdSense ads. AdSense may set third-party cookies for advertising purposes, including personalization based on your prior browsing if you have allowed Google to do so on other sites. You can review and manage Google's ad personalization settings in your Google account ad settings. EEA, UK, California, and other regional rights are explained in our full Privacy Policy.
Some "further reading" links lead to Amazon product pages and include our Amazon Associates affiliate identifier. If you click such a link and subsequently make a purchase on Amazon, we earn a small referral fee at no extra cost to you. The presence or absence of an affiliate identifier has no influence on our choice of book; books are selected on editorial merit.
We never receive payment, discount, or any other consideration from any author, publisher, or vendor for inclusion. No book has ever been added to the recommendation list because of a paid relationship.
Original event descriptions, the editorial selection of events, the organization of events into views and sections, the canvas layout system, and the source code are © 2025–2026 Cosmic History Timeline. All rights reserved.
Images displayed on this site are the work of their respective creators and are reproduced under the licenses indicated beneath each image. Most are Wikimedia Commons content available under one of the Creative Commons licenses or as Public Domain. Reuse of these images is governed by their original license, not by ours; click through to the Commons file page to see exact terms.
Wikipedia text linked from event popups is the work of its respective contributors and is licensed under CC BY-SA 3.0. Linking is permitted; reproduction is governed by Wikipedia's terms.
If you wish to reuse our original prose (event descriptions in Chinese or English) for non-commercial educational or research use, you may do so with attribution back to this site. For commercial use, please reach out via the contact link below.
For curious readers and prospective contributors:
events.json file with sections for axes, views, view groups, eras, era buttons, filter categories, and events. Loaded once on page open.generate-pages.py) that consumes events.json and emits one HTML file per view. Re-runs in about ten seconds; output is checked into the same git repository as the canvas.fetch-image-credits.py) and cached into events.json alongside each event. Re-fetchable on demand.fetch-relevance.py) calls a small language model with each event's existing description and the topic of the target page, and stores the per-event answer in events.json as relevant_views.generate-pages.py with hreflang annotations for the bilingual reading pages.indexedit.html) that mirrors the reader page but adds inline-edit affordances. The editor exports updates as JSON which can be merged back into the dataset via Python patch scripts.This site is maintained as a labor of love. If you have a correction (a wrong year, a missing event, a mistranslation), a suggestion (a country we should add, a topic we're missing), or just a question, the best place to reach us is via our Ko-fi page:
You can leave a comment, send a message, or, if you've found the site useful, leave a small tip. Tips help cover hosting and the time spent on data curation; everything beyond that goes back into adding more content. There is no commercial entity behind this; just one person and a long, slow weekend project.
Thank you for reading, and welcome to thirteen-point-eight billion years of history on a single canvas.