What Genderize does

Genderize answers one question: what gender is most likely behind a name? Send a full name, like Sarah Johnson, or just a first name like peter, and the answer comes back as a gender, a confidence score, and the number of records behind it.

Genderize looks the name up in a large dataset of real-world name records from around the world and reports what the data says. Some names are unambiguous; others lean one way with less certainty, and the confidence score tells you which is which.

What the response contains

Every Genderize response contains four fields.

`name`	The name from the request, echoed exactly as sent.
`gender`	"male" or "female". null when no prediction can be made.
`probability`	Confidence of the prediction, between 0 and 1.
`count`	Data points behind the prediction.

How Genderize reads names

Send the fullest form of the name you have. Every part of a name carries signal: the given name is the dominant signal for gender, and the surname can sharpen the prediction — or settle it. Icelandic surnames are a vivid example: Jónsdóttir is used exclusively by women, so Alex Jónsdóttir resolves as female even though Alex alone leans male.

Name parts are identified by evidence, not position. Tanaka Hiroshi and Hiroshi Tanaka produce the same prediction.
You can send names with titles, initials, or nicknames — Genderize handles them. Dr. J. Sarah Johnson reads as Sarah Johnson.
You can send accented or unaccented spellings — Genderize handles both José and Jose.

The name field in the response always echoes the input exactly as sent.

Country scoping

The link between a name and a gender varies by country. Kim is mostly female worldwide but overwhelmingly male in Denmark, and a prediction scoped to a country uses that country's data.

Scope a prediction by sending a two-letter country code, like US or DK, along with the name. Every surface supports it.