Biometric face match with calibrated scores
Compares the portrait on the ID to the selfie and returns a match score you can threshold. Scores are calibrated across skin tones, lighting, and camera quality.
≥ 85 verified60–85 review< 60 rejected
NewResidence permits now supported
Know exactly who is on the other side of the screen.
Veriface verifies government IDs and matches them to a live selfie in under ten seconds. One drop-in flow, one API call, a decision you can act on.
- SOC 2 Type II
- Images deleted after decision
- 190+ countries
- 8.4s
- Median decision
- 99.2%
- Face-match accuracy
- 0.3%
- False accept rate
Verification session
ver_k3d9x2q8
MRZ OK
Driver's license · front
Selfie · live
Scanning document28%
Reading MRZ, fonts and holograms
- Authentic97%
- Face match93%
- Livenesspass
Trusted by onboarding teams at
- Northwind Bank
- Lumen Health
- Orbit Mobility
- Fairmont Rentals
- Kestrel Pay
- Atlas Markets
- Harbor Insurance
- Pico Wallet
- Northwind Bank
- Lumen Health
- Orbit Mobility
- Fairmont Rentals
- Kestrel Pay
- Atlas Markets
- Harbor Insurance
- Pico Wallet
Everything in one call
Verification that holds up, without the integration sprawl
Document forensics, biometrics and liveness run together and return a single, explainable decision.
Decisions in seconds
Document and face analyses run in parallel.
p50p95
8.4s14.8s
Document authenticity
Layout, security features, MRZ checksums, expiry, and front/back consistency.
- Hologram
- Microprint
- MRZ
- Font kerning
- Expiry
- Tamper
Presentation attack detection
Flags screens, printouts, masks and swapped portraits before they reach your product.
Five document types
Licenses, state IDs, passports, passport cards and residence permits across 190+ countries.
Built for your stack
A single multipart POST returns structured JSON. Webhooks fire on every decision so you can pipe results into your review queue, CRM or onboarding flow.
Explore the APIverification.completed
Fires once per session
verification.needs_review
Routes to your reviewers
verification.decided
Manual outcome applied
How it works
Four steps from stranger to verified user
Drop the modal into any page, or call the API directly from your own capture UI. Either way you get the same structured result.
View integration guideChoose an ID type
The user picks a driver's license, state ID, passport, passport card, or residence permit.
Capture front, back, and selfie
Guided camera frames with automatic image downscaling keep uploads small and sharp.
Analyze and match
Two vision analyses run in parallel: document integrity and face comparison with liveness.
Get a decision
Receive verified, needs_review, or rejected with per-check results, scores, and extracted fields.
42M+
Verifications processed
Across fintech, mobility and marketplaces
99.97%
API uptime, trailing 12 months
Multi-region with automatic failover
0.3%
False accept rate
Measured on an independent benchmark
< 10s
Typical end-to-end session
Including capture and decision
Developer first
One request. One typed response.
No SDK lock-in and no session juggling. Send the images, receive a decision, per-check booleans, confidence scores, and the fields read off the document.
- Multipart upload, up to 8 MB per image
- Deterministic decision thresholds you can tune
- Every result visible in the dashboard review queue
TypeScript
Node
Python
3.9+
Go
1.21+
cURL
any
verify.ts
const form = new FormData()
form.set("documentType", "drivers_license")
form.set("front", frontFile)
form.set("back", backFile) // omit for passports
form.set("selfie", selfieFile)
form.set("userReference", "usr_8f31")
const res = await fetch("https://api.veriface.dev/v1/verify", {
method: "POST",
headers: { Authorization: "Bearer idv_live_..." },
body: form,
})
const result = await res.json()
// {
// id: "ver_k3d9...",
// status: "verified",
// scores: { authenticity: 0.97, faceMatch: 0.93 },
// checks: { faceMatch: true, selfieLive: true, ... },
// extracted: { fullName: "Jane Sample", ... }
// }What teams say
Fewer vendors, faster onboarding
We replaced a three-vendor stack with one call. Conversion through onboarding went up eleven points and our review queue shrank by half.
PNPriya Natarajan
Head of Risk, Kestrel Pay
The explainability is the thing. Every decision comes with the checks that drove it, so support can answer customers without engineering.
THTomás Herrera
VP Engineering, Orbit Mobility
Integration took an afternoon. The modal handled camera permissions, retakes and downscaling so we didn't have to.
MOMaya Okafor
Product Lead, Fairmont Rentals
Pricing
Pay for decisions, not seats
Volume discounts kick in automatically after 10,000 verifications a month.
Starter
$0
100 verifications / mo
- All five document types
- Face match and liveness
- Dashboard review queue
- Community support
Growth
$1.20
per verification
- Everything in Starter
- Webhooks and API keys
- Custom decision thresholds
- Priority support
Enterprise
Custom
volume pricing
- Dedicated review tooling
- Data residency options
- SLA and audit logs
- Solutions engineering
FAQ
Questions, answered
What happens to the images after verification?+
Images are processed in memory and discarded once a decision is returned. Only the structured result, scores and extracted fields are stored, and you can disable field retention per key.
Which documents are supported?+
Driver's licenses, state and national ID cards, passports, passport cards and residence permits from 190+ countries. Passports skip the back capture automatically.
Can we tune the decision thresholds?+
Yes. Growth and Enterprise plans can set per-key thresholds for face match and authenticity, and route borderline sessions to the review queue instead of rejecting.
How do manual reviews work?+
Sessions scored between your thresholds land in the dashboard queue. Reviewers see the checks, scores and extracted fields and approve or reject; a webhook fires with the final outcome.
Ship identity verification this afternoon
Create an API key, paste the modal into your onboarding flow, and watch sessions land in the dashboard.