Search companies with filters

Which lists: "any" for every Company list the caller can see, or an array of list ids. Membership is the UNION across the selected lists.

in — only companies who are a member of the selected list(s). notIn — exclude every company who is a member of the selected list(s); everyone else passes, including companies on none of your lists.

Number of the entity's loans in qualified affordable-housing tracts. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Share of the entity's tract-resolved loans in qualified census tracts for affordable-housing programs, 0-100 ({gte:25} = 27,209 originators). Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Mean Asian (non-Hispanic) share of the population in the tracts the entity lent in, 0-100. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Mean Black (non-Hispanic) share of the population in the tracts the entity lent in, 0-100. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Mean share of mortgage applications denied in the tracts the entity lent in, 0-100. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Mean median family income of the tracts the entity lent in, in whole dollars. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Mean Hispanic or Latino share of the population in the tracts the entity lent in, 0-100. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Mean median home value across the tracts the entity lent in, in whole dollars. A neighborhood statistic — not the entity's average loan size, which is avgLoanAmount. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Mean owner-occupied share of housing units in the tracts the entity lent in, 0-100. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Mean median household income of the tracts the entity lent in, in whole dollars (e.g. {lte:60000}). Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Mean tract income relative to its metro area, where 100 is parity — {lt:80} is the CRA low/moderate band, and values legitimately exceed 100 (observed up to 412). NOT a 0-100 share. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Bound the company's mean loan size for the selected period, derived from TOTAL production across ALL markets — NOT scoped to any geographic filter (city/state/zip/zipCode/geoPoint) in this request. For production INSIDE the geography this request names, read scopedVolume/scopedUnits on the same row, and sort by them to rank companies by it. There are two ways to get production for a geography: POST /companies/{nmlsId}/markets breaks ONE company's production out by market, and POST /loans/analytics/chart with measure: "volume", slice: "loanCompany" and a geographic filter answers the reverse — the companies that actually produced in that geography, each with the volume they produced there.

Mean HUD block-group low/moderate-income HOUSEHOLD share across the tracts the entity lent in, 0-100. A finer-grained measure than lowModIncomeTractPct, which counts whole tracts — not a duplicate of it. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Mean median age of the population in the tracts the entity lent in, in years. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Mean straight-line distance to the nearest college, in miles. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Mean straight-line distance from the properties the entity lent on to the nearest school, in miles. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Mean straight-line distance to the nearest place of worship, in miles. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Mean non-white share of the population in the tracts the entity lent in, 0-100 ({gte:50} = 63,370 originators). majorityMinorityTractPct is the per-tract-threshold form of the same question. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Mean population of the tracts the entity lent in. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Mean share of the population below the poverty line in the tracts the entity lent in, 0-100. Pass whole percents — {gte:0.2} means 0.2%. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Mean number of mortgage applications reported in the tracts the entity lent in, over the year. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Mean mortgage dollar volume reported in the tracts the entity lent in, over the year. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Mean number of mortgage originations reported in the tracts the entity lent in, over the year. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Mean White (non-Hispanic) share of the population in the tracts the entity lent in, 0-100. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Registered business address CITY — where the company is headquartered, NOT where it produced loans. This matches companies whose registered address is in this city; a company headquartered elsewhere is excluded no matter how much it lends here, and a company headquartered here is returned even if it produced nothing here. The row's volume/units/avgLoanAmount remain national totals, so a city filter combined with the default volume sort ranks the largest national producers that happen to be headquartered in this city. For production INSIDE the geography this request names, read scopedVolume/scopedUnits on the same row, and sort by them to rank companies by it. There are two ways to get production for a geography: POST /companies/{nmlsId}/markets breaks ONE company's production out by market, and POST /loans/analytics/chart with measure: "volume", slice: "loanCompany" and a geographic filter answers the reverse — the companies that actually produced in that geography, each with the volume they produced there.

Registered business address COUNTY — where the company is headquartered, NOT where it produced loans (same caveats as state; under locationBasis: "production" it selects companies that produced loans in the county instead). Accepts a county name (Travis, with a sibling state when the name exists in several states), a 5-digit FIPS code (48453) or a dictionary county id; an array matches any. The address carries no county of its own, so it matches when the geocoded point of the address lies inside the Census boundary of the county (a company with no geocoded address matches no county). An unrecognized county is a 400. For production INSIDE the geography this request names, read scopedVolume/scopedUnits on the same row, and sort by them to rank companies by it. There are two ways to get production for a geography: POST /companies/{nmlsId}/markets breaks ONE company's production out by market, and POST /loans/analytics/chart with measure: "volume", slice: "loanCompany" and a geographic filter answers the reverse — the companies that actually produced in that geography, each with the volume they produced there.

Number of the entity's loans in HUD-designated difficult development areas. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Share of the entity's tract-resolved loans in HUD-designated difficult development areas, 0-100 ({gte:25} = 93,397 originators). Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Number of the entity's loans in distressed-or-underserved tracts. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Share of the entity's tract-resolved loans in distressed-or-underserved nonmetropolitan middle-income tracts, 0-100 ({gte:25} = 11,773 originators). Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Number of the entity's loans in a FEMA special flood hazard area. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Share of the entity's tract-resolved loans in a FEMA special flood hazard area, 0-100 ({gte:25} = 5,212 originators). predominantFloodZone gives the specific zone code. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Radius search — { lat, lon, radius } (e.g. radius "15mi" or "24km") — around any office on the agent's record; with locationSource office the CURRENT office, production their transactions.

true: broker shops — non-bank companies that send more of the period's volume to OTHER lenders than they fund themselves. false: everyone else. Computed per period from each company's lender mix. Cannot be combined with mode: "or".

How many of the entity's loans in the selected period resolved to a neighborhood — the DENOMINATOR every Community Lending share is taken over. Pair it with a share filter ({gte:25}, 60,399 originators) so a 100% share off two loans does not read as a 100% share off two hundred. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Number of the entity's loans that landed in a low- or moderate-income tract, as a count rather than a share. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Share of the entity's tract-resolved loans that landed in a low- or moderate-income tract under the Community Reinvestment Act, 0-100. {gte:50} = 'at least half their book is LMI' (36,977 originators, 655 companies, 648 branches). Pass whole percents — {gte:0.5} means half of one percent and matches nearly everyone. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Number of the entity's loans in majority-minority tracts. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Share of the entity's tract-resolved loans in majority-minority tracts, 0-100 ({gte:50} = 65,654 originators). Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Registered business address METRO AREA (MSA / CBSA) — companies headquartered in ANY county of the metro (the address point inside the boundary of a member county, as for county); under locationBasis: "production", companies that produced loans in it. 5-digit CBSA code (12420) or metro name (Austin Metro); an array matches any. Member counties per the OMB July-2023 delineation (Connecticut's pre-2022 counties included; 41860 San Francisco Bay Area is the nine-county Bay Area). Unknown/ambiguous names are a 400 naming candidates.

DEPRECATED — fuzzy name matching is unreliable (name pollution silently returns unrelated companies). Resolve via instantSearch and filter by nmlsId (exact NMLS id) instead. Still functional for now.

Prefix name search: every word must start a word of the name ("rock mort" → Rocket Mortgage). No fuzzy/nickname matching; relevance-ranked unless sort is set. Max 100 chars / 6 words (emails, URLs, phones ignored), else 400.

Number of the entity's loans in Opportunity Zone tracts. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Share of the entity's tract-resolved loans in federally designated Opportunity Zone tracts, 0-100 ({gte:10} = 33,520 originators — this is a small-share filter, a 50% floor finds almost nobody). Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Registered business address CENSUS REGION / DIVISION (<region>-division-<n>, Census numbering 1–9) — companies headquartered in ANY state of it; under locationBasis: "production", companies that produced loans in it. An array matches any.

Bound the number of loan officers who joined the company during the selected period — individuals whose sponsorship with the company began inside the window. IMPORTANT — this counts LICENSING EVENTS, not confirmed employer moves. A loan officer holds a separate license record for every state they are sponsored in, and this counts each individual whose sponsorship began or ended inside the window, so a single added or lapsed state license registers as an arrival or a departure even when the person's employment did not change. Measured on large lenders: 13-24% of a 12-month rosterLeft are still sponsored by the same employer today, and 33-50% of a 12-month rosterJoined were already sponsored by it before the window began. Treat both as directional indicators of movement, not as headcounts. rosterJoined and rosterLeft are measured from the same source on the same window basis, so unlike the retired loLoss — which had no arrivals figure computed the same way — the two can legitimately be combined into a churn rate. Both also grow with the window, so a longer period reports more than a shorter one and the periods are comparable in that direction.

Bound the number of loan officers who left the company during the selected period — individuals whose sponsorship with the company ended inside the window. IMPORTANT — this counts LICENSING EVENTS, not confirmed employer moves. A loan officer holds a separate license record for every state they are sponsored in, and this counts each individual whose sponsorship began or ended inside the window, so a single added or lapsed state license registers as an arrival or a departure even when the person's employment did not change. Measured on large lenders: 13-24% of a 12-month rosterLeft are still sponsored by the same employer today, and 33-50% of a 12-month rosterJoined were already sponsored by it before the window began. Treat both as directional indicators of movement, not as headcounts. rosterJoined and rosterLeft are measured from the same source on the same window basis, so unlike the retired loLoss — which had no arrivals figure computed the same way — the two can legitimately be combined into a churn rate. Both also grow with the window, so a longer period reports more than a shorter one and the periods are comparable in that direction.

Mean remoteness of the tracts the entity lent in, on an ORDINAL 1-10 scale (1 = metropolitan core, 10 = most remote). This is a code, not a percent — {gte:7} selects the deepest-rural books (8,886 originators). Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Number of the entity's loans in rural tracts. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Share of the entity's tract-resolved loans in rural tracts, 0-100 ({gte:50} = 24,006 originators). ruralGradient is the finer, ordinal form. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Registered business address STATE — where the company is headquartered, NOT where it produced loans. This matches companies whose registered address is in this state; a company headquartered elsewhere is excluded no matter how much it lends here, and a company headquartered here is returned even if it produced nothing here. The row's volume/units/avgLoanAmount remain national totals, so a state filter combined with the default volume sort ranks the largest national producers that happen to be headquartered in this state — not the largest producers in it. For production INSIDE the geography this request names, read scopedVolume/scopedUnits on the same row, and sort by them to rank companies by it. There are two ways to get production for a geography: POST /companies/{nmlsId}/markets breaks ONE company's production out by market, and POST /loans/analytics/chart with measure: "volume", slice: "loanCompany" and a geographic filter answers the reverse — the companies that actually produced in that geography, each with the volume they produced there.

Bound the company's TOTAL loan count for the selected period across ALL markets — the same national total the response units returns, NOT scoped to any geographic filter (city/state/zip/zipCode/geoPoint) in this request, exactly as described on the volume filter. For production INSIDE the geography this request names, read scopedVolume/scopedUnits on the same row, and sort by them to rank companies by it. There are two ways to get production for a geography: POST /companies/{nmlsId}/markets breaks ONE company's production out by market, and POST /loans/analytics/chart with measure: "volume", slice: "loanCompany" and a geographic filter answers the reverse — the companies that actually produced in that geography, each with the volume they produced there.

Number of the entity's loans in USDA-eligible tracts. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Share of the entity's tract-resolved loans in tracts eligible for USDA rural housing programs, 0-100 ({gte:50} = 90,392 originators). Broader than ruralTractPct — eligibility reaches well past what is classified rural. Values describe the U.S. Census tract (neighborhood) as a whole — tract-level aggregate statistics, never an individual, household or applicant attribute. Percentages are 0-100.

Bound the company's TOTAL dollar production for the selected period across ALL markets — this is the same national total the response volume returns, and it is NOT scoped to any geographic filter (city/state/zip/zipCode/geoPoint) in this request. Combined with one of those filters it means 'national volume at least this large, AND headquartered in that place' — not 'produced this much there'. For production INSIDE the geography this request names, read scopedVolume/scopedUnits on the same row, and sort by them to rank companies by it. There are two ways to get production for a geography: POST /companies/{nmlsId}/markets breaks ONE company's production out by market, and POST /loans/analytics/chart with measure: "volume", slice: "loanCompany" and a geographic filter answers the reverse — the companies that actually produced in that geography, each with the volume they produced there.

Registered business address ZIP code — where the company is headquartered, NOT where it produced loans, with the same caveats as state and city: a company headquartered elsewhere is excluded however much it lends in this ZIP, and the row's volume/units/avgLoanAmount remain national totals. Canonical key (matches the response DTO); zipCode is an alias kept for cross-entity consistency with loans/sales. For production INSIDE the geography this request names, read scopedVolume/scopedUnits on the same row, and sort by them to rank companies by it. There are two ways to get production for a geography: POST /companies/{nmlsId}/markets breaks ONE company's production out by market, and POST /loans/analytics/chart with measure: "volume", slice: "loanCompany" and a geographic filter answers the reverse — the companies that actually produced in that geography, each with the volume they produced there.

Alias of zip, identical behavior — the registered business address ZIP code, i.e. where the company is headquartered, NOT where it produced loans. The row's volume/units/avgLoanAmount remain national totals. For production INSIDE the geography this request names, read scopedVolume/scopedUnits on the same row, and sort by them to rank companies by it. There are two ways to get production for a geography: POST /companies/{nmlsId}/markets breaks ONE company's production out by market, and POST /loans/analytics/chart with measure: "volume", slice: "loanCompany" and a geographic filter answers the reverse — the companies that actually produced in that geography, each with the volume they produced there.

Canonical lender name (the value from lender autocomplete). Resolved to a lender id server-side; pass id instead when you have one.

Label / typed search string

Lender-dictionary id — the id from instantSearch / listLenders (e.g. "mortgage_rocket" or "mortgage rocket"), and the same value the lender filter takes. PREFERRED: matched directly against the stored lender key with no name resolution. canonical is still accepted and resolved server-side.

Minimum percent_of_volume (0–100); 0/undefined = > 0%

Opaque cursor from previous response for next page

Page size (1-100, default 25)

volume (default) or units (loan count).

Upper bound, 0–100 percent. No bucket = 0% (passes).

Lower bound, 0–100 percent of the period book.

The loan product. Send this OR products, not both.

Several products, ANY of which may satisfy the bounds on its own (each type's share is read separately, never summed): {products:["fha","va","usda"], minShare:30} = FHA, VA or USDA is on its own at least 30% of the book. Send this OR product.

Dollar volume of THIS type in the period, {gte, lte} — e.g. {gte: 5000000} = at least $5M of FHA. Evaluated inside the same type bucket as the share bounds; a company with no loans of the type does not match.

volume (default) or units (loan count).

Upper bound, 0–100 percent. No bucket = 0% (passes).

Lower bound, 0–100 percent of the period book.

Dollar volume of THIS type in the period, {gte, lte} — e.g. {gte: 5000000} = at least $5M of FHA. Evaluated inside the same type bucket as the share bounds; a company with no loans of the type does not match.