The words the docs and the responses use, in alphabetical order. The shared parameters are defined in full on Requests and errors.
A pin: the one hexagon containing a point. On /places, the places in that hexagon; on /cells routes, that hexagon's row; on /boundaries, every boundary containing the point.
A rectangle as south,west,north,east, the way a map viewport asks. Latitude first.
The opposite sieve to within: keeps the hexagons where the nearest target is further than the time, or out of reach.
A country, nation, region, county, borough, ward, National Park or National Landscape. Ids are readable (gb-camden; gb-merton-abbey when two share a name) and stay the same while the boundary and its name do. Its hexagons: /boundaries/{id}?include=cells.
The kinds of boundary a country has (borough, ward, national_park…), listed on /catalog: the type filter on /boundaries, and, except National Parks and National Landscapes, a group_by value where a route offers one.
The hexagons you can reach within a travel time. The TravelTime API returns them with your own key; send them as cells.
The field a hexagon is always in: "cell": "89195da4d03ffff". On the geohash grid the field is geohash.
Exactly these hexagons: a comma-separated list of H3 ids in a URL, or in a POST body when too long.
A country id, gb. As a location, the whole country, on the searches that come back small.
On totals rows: how many people are within a travel time of a target, as covered_population and covered_share, without dropping any hexagon. Needs group_by.
The handle to the next page of a list. Pass next_cursor back as cursor with the same other parameters; null means you have everything.
1 is the most deprived tenth of areas in that nation, 10 the least. The King's Cross hexagon is decile 4.
Metres from near to each place, station, open space or road entry or exit in a list.
Country-wide files per dataset, listed on /catalog; none are published yet.
json (default) or geojson: a FeatureCollection with points for places and open spaces and hexagon outlines for /cells rows.
On places routes, geohash squares instead of hexagons: grid=geohash with precision 6 or 7.
How a /cells route totals what you sent: cell (default, one row per hexagon), a boundary type (one row per boundary your hexagons fall in) or all (one row for everything).
An H3 cell. The map is cut into them; resolution 9 (about 0.1 km², a city block or two) is the default, and each dataset lists the resolutions it has on /catalog.
Says what it is and where: pl_gb_ place, st_gb_ station, ln_gb_ line, os_gb_ open space, rd_gb_ road, ra_gb_ road entry or exit; boundaries are readable (gb-camden). Place ids change at each data update: store a place's name and coordinate, not its id. Station, line, boundary and open-space ids stay. Data updates
serves, parent, road, operator: a where of their own, needing no location.
Adds optional detail to a row: categories, groups, bands, open_spaces, cells, stops.
Picks neighbourhood blocks: noise, air_quality, flood, crime, deprivation, broadband.
Noise averaged over a day, with evening and night counted louder, in dB. The WHO's guideline for road traffic is 53 dB; the King's Cross hexagon's road noise is 66.7 dB, its loudest source 79.2.
Rows per page. A page size, not a cap.
Every list looks the same: {results, next_cursor}, a page at a time, plus sources where fields are sourced. The same shape as GET or POST.
On one place: the open datasets that list it, fsq (Foursquare), osm and overture in GB.
The only gap mechanism. A row's missing: [{field, reason}] says why a field has no value, e.g. {"field": "walk_s.post_office", "reason": "over_900_s"}. Nothing silently disappears.
Always a transport mode on transport routes: bus, coach, rail, underground, overground, ferry, cable_car; per country on /catalog.
The centre and radius (metres) of a circle.
Nearest first for near on places, stations, open spaces and roads; best match first for a name search; in the order sent for cells on /cells routes; largest first for at on /boundaries, otherwise by id; sort where offered.
The /cells routes: one row per hexagon, or totals with group_by. Things you can point at (places, stations, lines, open spaces, boundaries) have a list, one by id and /cells; facts about the land (demographics, neighbourhood, proximity) are /cells only.
Your own rules for custom scores: checks keyed by name, each with a field, min and/or max, weight (points if met), must (a dealbreaker) and label. One profile per request.
ratings.walk, .transit and .drive, 0–100, on /proximity/cells rows; null, with a missing reason, where a hexagon has no times.
The catchment of one start: the hexagons the TravelTime API says you can reach from a point within a travel time by one mode.
Hexagon size, 6–10, default 9; each step down is 7× bigger. Each dataset lists what it has on /catalog.
The rows of every list.
In every totals response: how each field was added up, by name, with one of sum, distinct, max, population_weighted_mean, household_weighted_mean, area_weighted_mean, ratio_of_sums, per_nation_population_share, none.
within and beyond: filters that keep or drop hexagons by walk or drive time. With country=gb they search the whole country.
On /open-spaces: sort=-area_m2, largest first.
One top-level map naming the dataset behind each sourced field, once per response rather than on every row; a value is a list when the rows draw on more than one. /catalog lists every source id per country as source_ids.
Rows that add up what you sent, per boundary or for everything, with a rollup block.
The Data API never calls TravelTime live. /proximity is a precomputed cache: walk and drive times from every inhabited hexagon, up to 15 minutes (900 s). For a longer journey or a commute, get the catchment from the TravelTime API with your own key and send it as cells.
Seconds (walk_s, drive_s), metres (radius_m, distance_m), square metres (area_m2), shares from 0 to 1 (no_car_share: 0.641). A name means one thing everywhere: lists keep the plain name (modes), counts end in _counts (mode_counts).
true when a government register (FHRS, CQC, NHS…) or TravelTime's own transport data confirms a place.
A sieve of travel.target:seconds conditions, comma-separated: travel is walk or drive; target a category, train_station, metro_station, bus_stop, transit_stop or brand:<slug>. Keeps only the hexagons that meet every condition.