Keyword Statistics
Cite-ready statistics from the wordradar suggest corpus: how many keywords exist, how many surface as brand-new each period, how the corpus grows day by day, and how it splits by lifecycle type and topic. Every number below comes from the same ledger that powers the API — free to reference, free to pull.
- CORPUS WORDS
- 28,129
- NEW WORDS (LATEST)
- 8,830
- PERIODS TRACKED
- 2
- LATEST PERIOD
- 2026-11-14
New-word pulse timeline
Words first seen in the corpus per period — the heartbeat of the ledger. Each bar is a corpus sweep; spikes mark days fresh phrases broke into Google suggest.
Period: 2026-11-13 → 2026-11-14 · First-seen words
Corpus composition · By lifecycle type
- 存量19,299 (68.6%)
- 新词8,830 (31.4%)
By topic direction
- AI5,419
- 工具1,465
- 游戏1,441
Type is recomputed on every sweep: a word is NEW only while its first period is the latest one; evergreens stay from day one; re-spiking terms count separately. Direction tags come from published term lists, not model guesses.
Latest cohort sample
A slice of the newest words — browse the full Top 20 boards on the live radar. Open live radar →
Data through · Sources: Hacker News
How these numbers are computed
First-seen means a keyword's first appearance in the multi-period suggest corpus — not a trends reading, not a volume estimate. A period is one corpus sweep; cohorts are immutable once recorded, while lifecycle types are recomputed each sweep as words age. All figures update daily with the latest bundle import, and the pulse endpoint serves the same numbers as JSON. Read the full methodology
Cite or pull these numbers
The pulse endpoint is free and unauthenticated — perfect for dashboards, reports and agent workflows. Per-keyword ledger data, verdicts and SERP difficulty use the metered API (1–2 credits).
WORDRADAR · OPEN KEYWORD STATISTICS · DAILY