Trustfull provides ready-to-use standard models for its products. Use model_id to choose a model when the endpoint supports model selection. For Trustfull Hub requests, model_id is required. Other product endpoints use the active model when you do not provide one.

Model name	Model ID	Description
Phone Number score	`phone_score`	Checks whether a phone number is real and active, identifies its line type, carrier, country, and linked online accounts, and assesses its risk.
Email Address score	`email_score`	Checks whether an email address is real and deliverable, enriches it with linked person, company, and online-account data, and assesses its risk.
IP Address score	`ip_score`	Checks an IP address's location and reputation and detects VPN, proxy, Tor, and hosting activity.
Consumer Onboarding	`digital_score`	Assesses consumer signups by checking whether the email address, phone number, name, and IP address consistently identify the same person. Disposable, unverified, or high-risk data lowers the score.
Business Onboarding	`business_onboarding`	Assesses business contacts using the Consumer Onboarding model, but excludes rules based on personal social-media accounts because business contacts may not have a personal social footprint.
iGaming	`igaming`	Assesses gaming and betting signups for multi-accounting and related risks. It flags email addresses or phone numbers associated with multiple identities, phone numbers with limited history, disposable contact details, and Tor usage.
Lead Validation	`lead_validation`	Checks whether a lead's contact details are valid and reachable. One issue, such as a disconnected, disposable, or incorrectly formatted phone number, lowers the score. Two issues place the lead in the lowest score band.
Mule Account	`mule_account`	Detects accounts that may have been opened to move money on behalf of others. Risk indicators include phone numbers or connections shared by multiple identities, remote desktop use, disposable phone numbers, Tor, mismatches between a person's name and contact details, and VPN, proxy, or hosting activity in another country. The model does not score requests without these indicators below the review band.
Return Fraud	`return_fraud`	Assesses orders and returns for signs of abuse. It flags phone numbers or connections associated with many email addresses, similar-looking email addresses linked to the same phone, disposable contact details, and email addresses with limited history. The model does not score requests without these indicators below the review band.
Domain score	`domain_score`	Checks whether a domain belongs to a real, operating business by analyzing its website, technical setup, company data, and reputation.
Document score	`document_score`	Checks uploaded payslips, tax documents, bank statements, and utility bills for editing or tampering, reads their content, and assesses their risk.
Account Takeover	`login_score`	Compares the device, browser, network, and behavior of a login attempt with the data saved at enrollment to detect possible account takeover.
Bot Detection	`bot_detection_score`	Detects automated or suspicious sessions using browser, device, network, and interaction signals such as headless-browser indicators and mouse activity.

FAQ

📘
Can I use a custom model?
Yes. Create the model in the Trustfull Dashboard. Once the model has been created, the dashboard displays its model ID.
Pass that ID as model_id in a request to a compatible Trustfull product. For Trustfull Hub, model_id is required.