What got filed,
where, and by whom.

This page reads U.S. utility patents granted in calendar 2024. The source is the USPTO Open Data Portal - PatentsView Granted Patent Disambiguated Data (product PVGPATDIS), which covers grants from 1976 to the present. Every count, share and CPC-section lean below is computed directly from those bulk tables.

How the numbers are built

real data The figures on this page are rebuilt from the PatentsView bulk tables - g_patent.tsv, g_assignee_disambiguated.tsv, g_cpc_current.tsv, and the disambiguated inventor + location tables - by scripts/build-data.ts. Counts are of granted, non-withdrawn utility patents. A short set of labels stays curated (anchor employers, sector examples, inventor field descriptors, the compare-page chip names); every number and every CPC lean is derived from the raw grants.

Top Assignees

Companies are ranked by U.S. utility patents granted in 2024. The year-over-year delta compares against 2023 grants for the same assignee. Patent ownership is messy - assignees can be subsidiaries, joint ventures, or rebranded entities - and PatentsView's disambiguation pass is what makes counts comparable. Counts are per disambiguated assignee; we only relabel each cluster to its recognizable family name (for example, one cluster is labeled "Samsung Display" in the source, but its 9,555-grant volume matches Samsung Electronics' U.S. output rather than the far smaller display arm, so it is shown as Samsung Electronics).

Geography

Metros are scored by inventor address, not assignee HQ. So Apple's patents land in San Jose-Sunnyvale-Santa Clara even though Apple is headquartered in Cupertino, because Cupertino is part of the same MSA and most named inventors live and work there. Multi-inventor patents are counted at the address of the first inventor. Each inventor's county is mapped to its metro (CBSA) using the Census 2023 delineation file.

The 18 metros shown are a curated selection of major U.S. invention centers, ranked here by their real 2024 grant counts - not the automatic national top 18. A few metros with real 2024 volume above the smallest shown here (notably Washington, Denver and Miami) are not on the map; the selection favors the recognizable innovation hubs.

The bubble map is drawn at build time: a geoAlbersUsa projection (with Alaska and Hawaii inset) over the Census state outlines from us-atlas, with each metro placed at a curator-authored centroid - an approximate [longitude, latitude] taken from the metro's principal city, not a population-weighted MSA centroid. Those coordinates are hand-authored reference points, precise enough to place a bubble on a national map but not survey-grade. Bubble area is proportional to the grant count. The map ships as static inline SVG with no client-side scripting.

Sectors

We use the top-level Cooperative Patent Classification (CPC) section codes - the broadest possible taxonomy. A patent counts toward each section it's classified under, so the percentages sum to slightly more than 100% in reality; here we normalize to the dominant classification per patent.

Inventors

Career counts are lifetime granted patents per disambiguated inventor. Inventor disambiguation is hard - "John Smith" matches dozens of people - so PatentsView uses ML to cluster patents likely belonging to the same person. The clusters can be wrong, especially for common names, so a very high count can overstate a single person. Each inventor's affiliation shown here is their most frequent assignee across those grants.

What's out of scope

Citation counts (which patents cite which) - the most-cited inventor list is more interesting than the most-prolific list, but citations require the g_us_patent_citation.tsv table, which is the biggest bulk file. Pre-grant publication data, plant patents, and reissues are out of scope here. Inventor career counts are lifetime disambiguated grants across all patent types, so a few prolific design-patent inventors surface alongside the utility leaders.