# Genderize.io

> Predict the gender most likely behind a name with Genderize.io, from a dataset of around a billion name records covering 249 countries. Gender one name or a whole list in a single step, read a name the way a particular country uses it, or compare a name across countries to see whether the answer changes. Every prediction comes with a confidence score and the number of records behind it, so a workflow can act on the ones it is sure about and leave the rest alone. Free for 2,500 names a month.

App · Contact Management · 3 actions. Web page: https://docs.lodol.com/docs/integrations/genderize

## Actions

### Compare a name across countries

Read one name in several countries at once and see whether they agree - Kim is female in the United States and male in Denmark. Use it to decide whether a name is safe to act on when you do not know where someone is from. The worldwide reading comes back too, as the baseline. Costs one name from your monthly plan per country, plus one for the worldwide reading.

**Inputs**

- Name (`name`, text, required): The name to read in each country.
- Countries (`countries`, list, required): The countries to compare, up to 10. One or more of: `AF` (Afghanistan), `AX` (Åland Islands), `AL` (Albania), `DZ` (Algeria), `AD` (Andorra), `AO` (Angola), `AI` (Anguilla), `AQ` (Antarctica), `AG` (Antigua & Barbuda), `AR` (Argentina), and 239 more.

**Outputs**

- `agreed_gender` (text): The gender all the countries agreed on. Empty when they disagree, or when none of them had an answer.
- `by_country` (list): One reading per country, in the order they were chosen.
- `name` (text): The name that was compared.
- `names_remaining` (number): How many names are left on your Genderize plan this month, as of this step. Empty if Genderize did not report it.
- `same_everywhere` (true/false): Whether every country that had an answer gave the same gender. False means the name reads differently depending on where someone is from.
- `worldwide` (object): The reading from Genderize's worldwide data, to compare the countries against.

**Example output**

```json
{
  "agreed_gender": null,
  "by_country": [
    {
      "confidence_percent": 94.0,
      "country": "US",
      "country_name": "United States",
      "gender": "female",
      "has_prediction": true,
      "name": "Kim",
      "probability": 0.94,
      "records": 196601
    },
    {
      "confidence_percent": 97.0,
      "country": "DK",
      "country_name": "Denmark",
      "gender": "male",
      "has_prediction": true,
      "name": "Kim",
      "probability": 0.97,
      "records": 196601
    }
  ],
  "name": "Kim",
  "names_remaining": 2494,
  "same_everywhere": false,
  "worldwide": {
    "confidence_percent": 75.0,
    "country": null,
    "country_name": null,
    "gender": "female",
    "has_prediction": true,
    "name": "Kim",
    "probability": 0.75,
    "records": 196601
  }
}
```

### Predict gender for a list of names

Predict the gender behind every name in a list, in one step - a column from a spreadsheet, the attendees of an event, a batch of new signups. Results come back in the order the names went in. Repeated names are looked up once, so only the names Genderize is actually asked about come off your monthly plan. Up to 500 names per step.

**Inputs**

- Names (`names`, list, required): The names to look up, one per entry. Full names are read most accurately. Blank entries are skipped.
- Country (`country`, text): Optional. Read the name the way it is used in one country - Kim is mostly female worldwide and mostly male in Denmark. Leave it empty to use Genderize's worldwide data. One of: `AF` (Afghanistan), `AX` (Åland Islands), `AL` (Albania), `DZ` (Algeria), `AD` (Andorra), `AO` (Angola), `AI` (Anguilla), `AQ` (Antarctica), `AG` (Antigua & Barbuda), `AR` (Argentina), and 239 more.

**Outputs**

- `count` (number): How many names were looked up.
- `names_looked_up` (number): How many names this step took off your monthly plan. Lower than the count when the list repeats a name.
- `names_remaining` (number): How many names are left on your Genderize plan this month, as of this step. Empty if Genderize did not report it.
- `predicted_count` (number): How many of them Genderize could name a gender for. The rest are not in its data.
- `predictions` (list): One prediction per name, in the order the names were given.

**Example output**

```json
{
  "count": 2,
  "names_looked_up": 2,
  "names_remaining": 2497,
  "predicted_count": 2,
  "predictions": [
    {
      "confidence_percent": 100.0,
      "country": null,
      "country_name": null,
      "gender": "male",
      "has_prediction": true,
      "name": "Peter",
      "probability": 1.0,
      "records": 1352696
    },
    {
      "confidence_percent": 95.0,
      "country": null,
      "country_name": null,
      "gender": "female",
      "has_prediction": true,
      "name": "Lois",
      "probability": 0.95,
      "records": 59503
    }
  ]
}
```

### Predict gender from a name

Predict the gender most likely behind one name, with how confident that is and how many people it is based on. Send a full name where you have one - the surname adds signal - or just a first name. Choose a country to read the name the way that country uses it. Costs one name from your monthly plan.

**Inputs**

- Name (`name`, text, required): The name to look up. A full name is read most accurately, and titles, initials and nicknames are handled - "Dr. J. Sarah Johnson" reads as Sarah Johnson.
- Country (`country`, text): Optional. Read the name the way it is used in one country - Kim is mostly female worldwide and mostly male in Denmark. Leave it empty to use Genderize's worldwide data. One of: `AF` (Afghanistan), `AX` (Åland Islands), `AL` (Albania), `DZ` (Algeria), `AD` (Andorra), `AO` (Angola), `AI` (Anguilla), `AQ` (Antarctica), `AG` (Antigua & Barbuda), `AR` (Argentina), and 239 more.

**Outputs**

- `confidence_percent` (number): The same confidence as a percentage, from 0 to 100, for text and for thresholds written in percent.
- `country` (text): The country the prediction was scoped to, as a two-letter code. Empty when the reading is worldwide.
- `country_name` (text): That country's name. Empty when the reading is worldwide.
- `gender` (text): The likelier gender behind the name, "male" or "female". Empty when Genderize has no data for the name.
- `has_prediction` (true/false): Whether Genderize could name a gender at all. False means the name is not in its data - not that the name is ambiguous.
- `name` (text): The name that was looked up, as you spelled it.
- `names_remaining` (number): How many names are left on your Genderize plan this month, as of this step. Empty if Genderize did not report it.
- `probability` (number): How confident the prediction is, from 0 to 1. A name used by both genders sits near 0.5.
- `records` (number): How many people in Genderize's data carry this name. A prediction backed by a handful of records is a weak one however high its confidence.

**Example output**

```json
{
  "confidence_percent": 100.0,
  "country": null,
  "country_name": null,
  "gender": "male",
  "has_prediction": true,
  "name": "Peter",
  "names_remaining": 2499,
  "probability": 1.0,
  "records": 1352696
}
```