Search

GET /sanctions/search finds records by text in the caption. Every parameter, the filters, sorting, paging, and what the contains match means for you.

GET /sanctions/search

search returns a page of records whose caption contains your text. It is the endpoint behind a list view: a user types a name, reads the hits and opens one.

For a screening decision, use POST /sanctions/match instead. Read Search or match.

Full field list: searchSanctions in the reference.

Parameters

All parameters go in the query string. All values are strings.

Parameter	Values	Default	What it does
`query`	text	none	The text to find in the caption. It needs at least one token with three or more letters or digits. Without it you get the whole (filtered) list.
`type`	`sanction`, `pep`	`sanction`	Which body of data to read.
`page`	integer from 1	`1`	The page to return.
`per_page`	integer from 1 to 100	`50`	Records per page.
`dataset`	comma-separated names	none	Keep records from these source datasets. Names come from `GET /sanctions/datasets`.
`schema`	comma-separated names	none	Keep records of these kinds, for example `Person,Company`. Names come from `GET /sanctions/schemas`.
`country`	comma-separated codes	none	Keep records with one of these countries.
`sort_method`	`relevance`, `alpha`, `date`	`relevance` with `query`, else `date`	The sort order. See below.
`sort_order`	`asc`, `desc`	`desc`	The direction for `alpha` and `date`.
`target`	`true`, `false`	`true`	`true` returns only entities that are themselves listed. Send `false` for the supporting records.
`whole_word`	`true`, `false`	`false`	`true` requires your text to stand as a whole word in the caption.
`from_date`	`YYYY-MM-DD`	none	Keep records changed on or after this date.
`to_date`	`YYYY-MM-DD`	none	Keep records changed on or before this date.

An unknown parameter is an error, not something the API ignores. limit=5 returns 400 unknown_parameter. Read Errors.

The response

{
  "results": [
    {
      "id": "es-mc-85d50463dadae4256272c2014ee57ba9ce55b92e",
      "caption": "LOURDES GULINA CANEDA",
      "schema": "Person",
      "datasets": ["es_mayors_councillors", "ann_pep_positions"],
      "last_change": "2025-11-05T14:35:45",
      "properties": {
        "country": ["es"],
        "name": ["LOURDES GULINA CANEDA"]
      }
    }
  ],
  "total": 4,
  "page": 1,
  "per_page": 2,
  "total_pages": 2
}

Field	What it is
`results`	The records on this page.
`total`	Every record that fits the query, over all pages.
`page`	The page you got back.
`per_page`	The page size used.
`total_pages`	The number of pages at this page size.

A search result is a short entity: id, caption, schema, datasets, last_change and a part of properties. Call GET /sanctions/entities/{id} with the id for the full record.

Sorting

`sort_method`	Sorts by	Notes
`relevance`	How well the caption fits `query`	The default when you send `query`. A whole-word hit ranks above a hit inside a word. For a query of two or more words, a hit with the same word order ranks highest. `sort_order` has no effect.
`alpha`	`caption`	A to Z with `sort_order=asc`.
`date`	`last_change`	Newest first with the default `sort_order=desc`. The default when you send no `query`.

relevance needs a query. Without one, the API sorts by date.

Paging

Read total_pages, then walk the pages with page. The maximum page size is 100.

curl -s -G "https://api.linkinlegal.com/v1/sanctions/search" \
  -H "Authorization: Bearer $LINKINLEGAL_API_KEY" \
  --data-urlencode "query=Gulina" \
  --data-urlencode "per_page=100" \
  --data-urlencode "page=2"

Both relevance and date break ties on the entity id, so a record does not jump between pages while you walk them.

Filters

Combine the filters. They all narrow the same list.

# Companies from the EU financial sanctions file, registered in Russia.
curl -s -G "https://api.linkinlegal.com/v1/sanctions/search" \
  -H "Authorization: Bearer $LINKINLEGAL_API_KEY" \
  --data-urlencode "schema=Company,Organization" \
  --data-urlencode "dataset=eu_fsf" \
  --data-urlencode "country=ru" \
  --data-urlencode "per_page=25"

# Everything that changed since 1 September 2026, newest first.
curl -s -G "https://api.linkinlegal.com/v1/sanctions/search" \
  -H "Authorization: Bearer $LINKINLEGAL_API_KEY" \
  --data-urlencode "from_date=2026-09-01" \
  --data-urlencode "sort_method=date"

from_date and to_date read the last_change of the record. Read Data freshness before you build a "what is new" job on them.

Ask GET /sanctions/datasets, GET /sanctions/schemas and GET /sanctions/countries for the values these filters accept. A dataset name that does not exist is not an error: it returns an empty list.

What "contains" means for you

The match is a plain case-insensitive "contains" on the caption. Four consequences:

Your text can sit inside a longer word. query=Gulina returns Luopu County Gulina Muhan Clothing Co., Ltd. Send whole_word=true to require word boundaries.
The word order must be the same. The caption Vladimir Putin is found by query=Vladimir Putin, and not by query=Putin Vladimir.
Only the caption is read. Aliases, former names and other spellings are in properties, and search does not look there.
The script must be the same. A caption in Cyrillic is not found with Latin text, and the other way round.

# Whole word: finds "Vladimir Putin", not "Computing".
curl -s -G "https://api.linkinlegal.com/v1/sanctions/search" \
  -H "Authorization: Bearer $LINKINLEGAL_API_KEY" \
  --data-urlencode "query=Putin" \
  --data-urlencode "whole_word=true"

Because of points 2, 3 and 4, an empty search result does not mean the subject is clean. Use match for that answer: it compares every name and alias, it accepts a different word order, it tolerates typos, and it reads Cyrillic and Greek names in Latin letters.

Errors

Status	Code	Cause
400	`query_too_short`	`query` holds no token with three or more letters or digits.
400	`unknown_parameter`	A parameter name that does not exist.
400	`invalid_request`	A value out of range, for example `per_page=500` or `page=0`.

Read Errors for the rest.