Build with Fingerly. Decide with evidence.
Identify the device behind every session, score how suspicious it is, and act on it from your own server. Everything you need, from the first key to production.
import { load } from '@fingerly/web-js'
const fingerly = await load({ apiKey: 'fly_pk_us_production_…' })
// At the moment that matters
const { requestId } = await fingerly.identify({ tag: 'checkout:8412' })
Every surface. One response.
Browsers, iPhones and Android phones return the same fields. Server SDKs read them back and verify webhooks.
Choosing an SDKWeb
Understand the model. Then tune it.
The ideas every integration is built on. Each page is short, and says exactly what the API does.
Visitor identification
What a visitor ID is, and what confidence means.
Suspect score
A sum of weights you can read, explain and change.
Signals reference
Every signal group and its default weight.
Risk weights
Tune signals and the threshold without redeploying.
API keys and environments
Public, secret and proxy keys; free test traffic.
Regions and data residency
Where your visitors’ data lives.
Start from the fraud. Work back to the code.
End-to-end recipes: where to identify, what to tag, what your server checks, and a policy to start from.
Account takeover
Known devices through, new and risky ones challenged.
Credential stuffing
Stop scripted logins, counted per device.
Sign-up abuse
Limit how many accounts one device creates.
Promotion abuse
One redemption per device, however many accounts.
Payment fraud
Review risky orders and stop card testing.
Migrate from FingerprintJS Pro
Map get(), Smart Signals and webhooks.
A small API. Documented exactly.
Client, server and management APIs, and five webhook events. Every field, header and error, with examples beside it.
API overviewOpen the dashboard
Create keys, add webhook endpoints, tune weights and explore every identification.
Go to the dashboardTalk to the team
Questions about an integration, a rate limit or your data. Try troubleshooting first, and include a request ID if you have one.
support@fingerly.ioUse it with AI tools
Every page has a Markdown twin at its URL plus .md, and the whole site is indexed for language models.