How well does the API handle gamer tags?

Tags that contain a recognizable name often return a result. Many handles do not, and the API returns unknown with a reason such as no_name_candidate. AI mode can help, at a higher credit cost.

Can predicted gender be used for matchmaking or lobbies?

No. Predictions are estimates that are wrong for some people and must not shape matchmaking, moderation or rewards. Use them only for aggregate audience research.

Can we combine forceToGenderize with ai_mode off?

No. forceToGenderize cannot be combined with ai_mode off or always, and the API returns a 400 error. Requests rejected with 400 are not charged.

Gender prediction · Gaming

Gender estimates from gamer tags for player research

Many players never give you more than a gamer tag. The Finowl gender API can estimate how names and some handles are typically used, and an optional AI mode reads nicknames and handles that contain no real name. Research teams can use those estimates, in aggregate, to describe the audience for a mode, region or marketing beat when survey data is missing.

Handles are noisy, so expect many unknown results and treat them as a normal outcome. Predictions must stay out of matchmaking, moderation, rewards, pricing and anything that changes how a player is treated. If your game offers pronouns or character options, let players choose freely and change them whenever they like.

Try it in the sandbox About gender

USE CASES

Where it fits in gaming

Mode and region audience snapshots
Estimate the aggregate composition of players in a new mode or region to support research reports, keeping unknown results visible so teams understand how much of the audience is covered.
Playtest panel balance
Check whether an opt-in playtest panel appears to reflect your broader audience in aggregate, then widen recruitment channels if it looks narrow, instead of selecting individual players.
Marketing beat reporting
After a trailer or seasonal event, estimate the aggregate audience of new signups to understand reach, combining predictions with your existing channel and region data.
Handle-only enrichment with AI
For tags that contain no real name, forceToGenderize lets AI read nicknames after the dataset lookup, at 3 credits per request, with a model-reported confidence you should interpret cautiously.

HOW TO INTEGRATE

The workflow

1
Export gamer tags, or first names where players have provided them, into a research dataset separate from live services.
2
Call POST /v1/gender/username for each tag, optionally with forceToGenderize for handles without a real name.
3
Record gender, confidence, confidence_kind and source so AI-based results can be reported separately.
4
Aggregate results by mode or region with unknowns shown as their own category.
5
Keep the field out of matchmaking, moderation, store and reward systems.

POST /v1/gender/username

curl https://api.finowl.co/v1/gender/username \
  -H "Authorization: Bearer $FINOWL_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "username": "nightowl_jess",
  "forceToGenderize": true
}'

forceToGenderize tries the dataset first, then AI reads the handle. It costs 3 credits and AI results return source ai with confidence_kind model_reported.

GOOD PRACTICE

Before you ship

Never use predictions for matchmaking, moderation, chat filtering, rewards or pricing. These shape how players are treated and must not depend on a statistical estimate.
Game audiences include minors. Check age-related obligations before processing player data, keep analysis aggregate and mention gender inference in your privacy notice.
Many handles carry no gender signal. Report unknowns honestly, and use self-selected pronoun or character options in game rather than inferring them from a tag.

Read the API reference for every field and error code. Each lookup costs 1 credit (3 when it requests AI); see pricing.

A FEW GOOD QUESTIONS

Common questions