Software Development Kits
Biometric Algorithms, Ready to Embed.
id3's SDKs deliver NIST-validated face recognition, fingerprint identification, document verification, and on-card biometrics as native libraries — for servers, mobile devices, and microcontrollers. One consistent API surface, every platform.
Available SDKs
Six production-ready libraries covering every biometric modality — from cloud servers to EAL6+ secure elements.
Face SDK
Cross-Platform Face Recognition
Full-featured face recognition library for servers, desktops, and mobile devices. NIST FRTE validated accuracy. Covers detection, comparison, 1:N identification, liveness, and portrait quality.
- Detection, comparison, 1:N identification
- Liveness & portrait quality
- Windows · Linux · macOS · Android · iOS
- NIST FRTE evaluated
Finger SDK
Fingerprint Identification
Standardized, sensor-independent fingerprint recognition with NIST MINEX III validation. Supports ISO 19794-2 and ANSI 378 template formats for full interoperability.
- Extraction & 1:N matching
- Sensor-independent
- ISO 19794-2 / ANSI 378
- NIST MINEX III evaluated
Document SDK
Automated Document Verification
Automatic identification, MRZ reading, OCR extraction, barcode decoding, and security element checks for passports, ID cards, visas, and driving licences from 200+ countries.
- MRZ, OCR, barcode, NFC
- 200+ country documents
- Tampering & forgery detection
- Face extraction from document
MicroFace SDK
Embedded Face Recognition
Face recognition library optimized for microcontrollers and edge devices. Runs on Cortex-M55/M85, i.MX 8M Plus, and STM32N6 — sub-100 ms inference with <1 MB footprint.
- Cortex-M55/M85 · i.MX 8M · STM32N6
- Sub-100 ms inference
- NPU/DSP accelerated
- <1 MB memory footprint
MicroFinger SDK
Embedded Fingerprint Recognition
NIST MINEX III fingerprint recognition for microcontrollers. The same algorithm quality as the full Finger SDK, optimized to run on resource-constrained hardware without compromise.
- NIST MINEX III algorithms
- Microcontroller optimized
- Sensor-independent capture
- Compact template format
Match-on-Card SDK
On-Card Biometric Authentication
Biometric matching entirely on the secure microcontroller of a smart card — no template ever leaves the card. EAL6+ certified, <10 kB footprint, MINEX III algorithms.
- Zero biometric data off-card
- EAL6+ certified
- <10 kB footprint
- NIST MINEX III algorithms
Why Choose an SDK
A native library gives you full control over the biometric pipeline — on your hardware, in your process, with no external dependency at runtime.
Data Stays On-Device
All biometric processing happens locally — no biometric data is sent over a network. Ideal for air-gapped environments, regulated industries, and privacy-first architectures.
No Latency Penalty
Native execution means no network round-trips. Sub-100 ms face comparison, real-time fingerprint extraction — performance limited only by the host hardware.
Works Offline
Zero connectivity requirement at runtime. SDKs operate in fully disconnected environments — border posts, ATMs, access control systems, and field devices all work without internet.
Multi-Language APIs
C, C++, C#, Python, Java, Kotlin, Swift, Dart — native bindings for every major language. Ready-made samples for Windows, .NET, Flutter, Android, and iOS.
NIST Validated Accuracy
Face SDK ranked in NIST FRTE. Finger SDK ranked in NIST MINEX III. Independent government benchmarks — not marketing claims — validate the accuracy you'll get in production.
From Cloud to Microcontroller
The same algorithms run on Linux servers, Android phones, and ARM Cortex-M microcontrollers. Scale from a cloud ABIS to an embedded access terminal without changing vendors.
Platform & Resources
One consistent API across every operating system, language, and hardware class — from cloud servers to credit-card-sized secure elements.
Operating Systems
Languages
Hardware Targets
Frameworks & Samples
Ready to integrate? Explore documentation, evaluation licenses, and code samples on our developer portal.
Developer Portal ↗