SmrAI Logo
RESTFUL & GRPC NEURAL ENDPOINTS • 14MS ULTRA-LOW LATENCY • SDK V2.4

Developer API & SDK — Multi-Modal AI Detection API

Integrate world-class AI detection, humanization, acoustic voice forensics, and deep plagiarism verification directly into your applications, enterprise LMS, or CI/CD automated review pipelines.

verified 99.995% SLA Cluster
memory Zero-Retention Ingestion Engine
encrypted TLS 1.3 & mTLS gRPC
api.smrai.io/v2 Keys & Quota ↓
Endpoint Inventory v2.4.1 LIVE
POST /v2/detect/multimodal
application/json

Conducts token-level neural probability inference across multi-turn text, code blocks, and embedded structural markdown. Returns synthetic confidence, LLM architecture attribution fingerprints, burstiness entropy, and sentence-by-sentence watermarking telemetry.

Required Headers
Authorization Bearer smr_live_*** Content-Type application/json X-Inference-Mode deep-forensic | fast
Schema Parameters
content string • required
Raw plain text, markdown, or code excerpt to evaluate (up to 250,000 chars per batch).
granularity string • default "sentence"
Inspection resolution: token, sentence, or paragraph.
include_fingerprint boolean • default true
Identify likely generating model family (GPT-4o, Claude 3.5 Sonnet, Gemini 1.5, Llama 3).
verified_user
Zero Storage Ingestion Policy Transmitted payloads are processed purely in ephemeral RAM and never indexed, stored, or reused for LLM model training.
Live Server Response HTTP 200 OK 14ms Latency
us-east-neural-1
Diagnostic Verdict: Live System Operational
Rate Limit: 1198 / 1200 RPM | Zero Storage: Ephemeral RAM
Credentials & Telemetry Usage

Active API Credentials & Live Quota

Manage production keys, monitor consumption rates, and bind live webhook dispatch targets.

Production Live Secret READ / WRITE
smr_live_948a31e87bb61099fd84
Created: Feb 28, 2025 • Never expires
Active TLS mTLS bind Rotate →
Monthly Inference Quota 64.2%
64,200 / 100,000 req
Resets in 18 days Upgrade Tier →
Asynchronous Webhook ACTIVE
Dispatches on batch completed & high synthetic alerts.
send HMAC SHA-256 Verified
terminal
Python SDK v2.4.0 • PyPI

Native async clients with automatic retries, streaming token analysis, and Pandas batch dataframe hooks.

pip install smrai-sdk content_copy
code_blocks
Node.js / TS v2.4.1 • npm

Full TypeScript types, Web-stream compatible for Next.js Edge, Remix, Cloudflare Workers, and Bun.

npm install @smrai/client content_copy
verified
SOC2 Type II & FERPA Compliance Guaranteed

Guaranteed zero-data retention on all standard API tiers. Dedicated VPC instances available with HIPAA BAA execution.

policy Download Security Whitepaper →
Telemetry Flow Pipeline

Ultra-Low Latency Inference Network

Requests are routed instantaneously to the closest inference edge cluster, running quantized entropy kernels in memory without cold starts.

Client App / LMS REST / gRPC TLS 1.3 SmrAI Edge Gateway Token Sanitizer & Auth (2ms) Zero Disk Buffer RAM Neural Forensic Clust. 12ms Diagnostic Verdict
Global Anycast Ingestion
Sub-15ms Roundtrip P95
Multi-Region Failover