What changes between these APIs?
Both services document name, email and username workflows. Important differences include AI defaults, batch structure and confidence semantics. NameGender best_guess and GenderAPI.io forceToGenderize describe different behaviors and should not be mapped automatically.
| Decision point | GenderAPI.io V2 | NameGender.com | Sources |
|---|---|---|---|
| Inputs and authentication | Names, email addresses and usernames use type and value on /api/v2/gender. Authorization: Bearer with a GenderAPI.io key. Query keys are available for simple GET examples; prefer headers in applications. | Separate /api/v1/gender, /gender/email and /gender/username routes; Bearer or X-Api-Key. URL keys are rejected. | [1][2][14] |
| Batch requests | Up to 50 keyed items or 10 IP-trial items. Each item can have its own type, country and options. | Up to 100 values per /api/v1/gender/bulk request; one input type and country for the whole batch. | [5][9] |
| Confidence and unknowns | data.confidence is 0–1 or null. confidence_kind distinguishes stored-count frequency from an AI score; sample_count is null for AI. data.result_status: unknown; data.gender and data.confidence are JSON null. Inspect data.reason. | probability: 0–100; confidence: a category such as high or unknown; sample_size: an integer. gender can be null. | [3][12] |
| AI fallback | Explicit off, fallback or always. Singles default to fallback; batches default to off. Optional forceToGenderize uses the dataset first, then nickname-aware AI. | ai_fallback defaults to false and requires account consent. The reference describes 1 extra credit when AI is used. | [4][14] |
| Nickname options | forceToGenderize applies to all three input types and can infer an alias association even when data.name is null. | best_guess lowers the name-evidence threshold. For email/username it acts on an extracted first name. | [4][11][10][15] |
| Unknown-result billing | Ordinary successful lookups cost 1 credit, even unknowns and fallback AI. Always-AI costs 2. forceToGenderize costs 1 for a resolved dataset result or 2 when AI runs. | An ordinary successful unknown still costs 1 credit. Read credits_charged and credits_remaining. | [6][11][12] |
| Payment models | Monthly subscriptions and one-time packages; see current plan terms. | One-time packages plus monthly/yearly subscriptions. Subscription credits reset monthly; one-time credits persist until used. | [8][13] |
Before you switch an integration
Map the endpoint-specific name, email or username field to type and value. Rework bulk names into items, respecting the smaller V2 limit. Use meta.usage.charged_credits and remaining_credits for usage. Select options.ai_mode explicitly: NameGender’s documented AI default differs from a GenderAPI.io single request.
Dividing a NameGender probability by 100 only changes its units. It does not make it statistically equivalent to GenderAPI.io confidence. NameGender confidence is categorical; GenderAPI.io confidence is numeric and confidence_kind describes its basis. Review both your parser and thresholds.
Evaluate the workflow you will deploy
Choose a representative sample you are permitted to process and keep the original inputs, preprocessing and country context fixed. Evaluate name, email and username workflows separately. Record provider, endpoint, date, configuration and whether AI was used.
Measure coverage (answered / eligible), accuracy among answered results (correct / answered), and correct results across all eligible inputs (correct / eligible). Report unknowns, errors, costs and latency separately. A threshold should be selected on validation data and then checked on a held-out sample.
GenderAPI.io does not publish an independent reference-labelled accuracy benchmark here. Database counts, contract tests and one provider's published study cannot establish which service will work best on your data. An inferred association does not establish a person's gender identity.
Sources:[7]
Frequently asked questions
Is best_guess the same as forceToGenderize?
No. NameGender documents a lower evidence threshold, with email and username behavior tied to an extracted name. GenderAPI.io nickname mode can attempt an alias association without a real-name result. Neither option guarantees a non-null or correct answer.
Sources and review dates
Use the linked documentation to confirm current terms and behavior before choosing a provider.
- GenderAPI.io V2 request parametershttps://www.genderapi.io/docs/v2/request-parametersReviewed
- GenderAPI.io V2 authenticationhttps://www.genderapi.io/docs/v2/authenticationReviewed
- GenderAPI.io V2 responseshttps://www.genderapi.io/docs/v2/responsesReviewed
- GenderAPI.io V2 AI optionshttps://www.genderapi.io/docs/v2/ai-optionsReviewed
- GenderAPI.io V2 batch guidehttps://www.genderapi.io/docs/v2/batchReviewed
- GenderAPI.io V2 credits and usagehttps://www.genderapi.io/docs/v2/credits-and-usageReviewed
- GenderAPI.io accuracy methodologyhttps://www.genderapi.io/accuracy-methodologyReviewed
- GenderAPI.io pricinghttps://www.genderapi.io/priceReviewed
- NameGender API: Bulk requesthttps://namegender.com/docs/ai/bulk.mdReviewed
- NameGender API: Gender from an emailhttps://namegender.com/docs/ai/email.mdReviewed
- NameGender API: Gender from a namehttps://namegender.com/docs/ai/name.mdReviewed
- NameGender API OpenAPI 3.1 contract, info.version 1.0.0https://namegender.com/openapi.jsonReviewed
- Pricing — NameGenderhttps://namegender.com/pricingReviewed
- REST API Reference — NameGenderhttps://namegender.com/docsReviewed
- NameGender API: Gender from a usernamehttps://namegender.com/docs/ai/username.mdReviewed