How Hacker Trends works
A Google-Trends-style explorer for 18 years of Hacker News, built on Upstash Redis Search.
What it does
Type any topic, tool, company, or person and Hacker Trends charts how often it has come up on Hacker News, month by month, since 2007. Overlay several terms to watch their traction rise and fall against each other — the same way you would compare search interest on Google Trends, but for the stories and comments that shaped the tech industry’s conversation. Open the tool →
How to use it
- Search a single term to see its mention-over-time line and the top stories behind it.
- Compare up to five terms at once — each gets its own colored line on a shared axis.
- Zoom to a date range by selecting months on the chart to drill into a specific spike.
- Filter the results by author or by type (story vs comment), and sort by relevance, points, comments, or recency.
- Share any view — the full state lives in the URL, so a link reproduces the exact chart and filters.
What’s under the hood
About 45 million Hacker News items from 2007–2026 are stored as plain Redis hashes and indexed with Upstash Redis Search. Every chart is computed live, with no separate analytics warehouse:
- The trend line is a
SEARCH.AGGREGATEdate-histogram — one monthly bucket per point — filtered to items that actually mention the term. - The stories list is a
SEARCH.QUERYfull-text search, ranked by a blend of BM25 relevance plus upvotes and comment count so the genuinely-discussed threads rise to the top.
Popular comparisons
Some of the rivalries and successions the data tells best:
- vercel vs cloudflare
- openai vs anthropic
- amd vs nvidia
- scala vs swift vs kotlin
- angular vs vue vs svelte
- mysql vs postgres
- tensorflow vs pytorch vs jax
- webpack vs vite
Or jump straight to a single term, e.g. chatgpt, rust, bitcoin, kubernetes.
Frequently asked questions
What is Hacker Trends?
Hacker Trends is a Google-Trends-style explorer for Hacker News. You type any topic, tool, company, or person and it charts how often that term has appeared in Hacker News posts and comments each month, so you can see when interest rose, peaked, and faded. You can overlay several terms on one chart to compare them.
Where does the data come from?
It indexes roughly 45 million Hacker News items — stories and comments — spanning 2007 to 2026. Each item's title, text, author, type, timestamp, score, and comment count are stored as a Redis hash and indexed for full-text search.
How far back does the data go?
The index covers Hacker News from 2007 through 2026 — about 18 years of the front page and its comment threads.
How is a term's popularity measured?
Each point on the line is the number of Hacker News posts and comments in that month whose title or body mentions the term — an honest, exact mention count, not a fuzzy or weighted score. The chart is a live date-histogram computed at query time.
What powers Hacker Trends?
It is built on Upstash Redis Search. The trend lines come from SEARCH.AGGREGATE date-histogram queries, and the list of stories behind each line comes from SEARCH.QUERY full-text search — both running directly against Upstash Redis with no separate analytics database.
How are the top stories ranked?
Relevance ranking blends the text-match (BM25) score with the story's upvotes and comment count, so genuinely discussed and upvoted threads surface ahead of incidental mentions. You can also sort purely by points, by comment count, or by recency.
Is Hacker Trends free to use?
Yes. It is a free, public demo built to show what Upstash Redis Search can do on a real, large dataset.