Catch API abuse
before it becomes a breach
Boing is a production-ready platform for monitoring APIs and user accounts. It detects suspicious activity across four layers of analysis and sends real-time alerts to your team by email, webhook, or Slack — all self-hosted, all free.
Everything you need to watch your APIs, nothing you have to pay for
Boing pairs a fast ingestion pipeline with a detection stack built for real production traffic — from a single self-hosted deployment.
Real-time monitoring
A WebSocket-based live activity stream shows every request the moment it hits your API — no polling, no delay.
Multi-layer detection
Rule-based, statistical, ML (Isolation Forest), and optional local LLM analysis work together so nothing slips through on a single blind spot.
Alerts where you already work
Email over SMTP, webhooks into Slack, and in-app notifications keep the right people in the loop the moment something looks wrong.
Security by default
Encrypted API secrets, JWT auth, role-based access, and rate limiting protect the monitor as carefully as the APIs it watches.
A dashboard that stays out of your way
A React + Vite dashboard surfaces live metrics and visualizations without needing a second tool to make sense of them.
Free and open source
No paid APIs, no metered detection calls. Boing runs on local models you control, under an MIT license.
From first request to first alert in three steps
No agents to run, no sidecars to deploy — just a middleware call and a running instance.
Drop in the middleware
Add a few lines to your Flask, FastAPI, or Express app. Every request's method, endpoint, headers, status, and latency streams straight to your Boing instance.
requests.post(BOING_URL, json={
"api_key": BOING_API_KEY,
"method": request.method,
"endpoint": request.path,
"status_code": response.status_code,
})The pipeline does the thinking
Every event passes through rules, statistics, machine learning, and — if you enable it — a local LLM, each layer catching what the last one missed.
Rules → Stats → ML (Isolation Forest) → LLM
z-score > 3.0 → flagged
rate > 100/min → throttledYour team gets the alert
A flagged event becomes a notification in the dashboard, an email, or a Slack message — with enough context to act on immediately, not just react to.
ALERT ip=185.220.101.7
type=rate_limit action=blocked
sent → email · slack · dashboardBuilt for teams who'd rather own their monitoring than rent it
A self-hosted stack means no per-request billing, no vendor lock-in, and no shipping your traffic to a third party to find out if it's suspicious.
4
Detection layers, from rules to a local LLM
<1s
Typical time from anomaly to alert
$0
Paid APIs required to run detection
100%
Self-hosted — your data never leaves your infra
Simple plans, or pay for exactly what you send us
Boing itself is open source and free to self-host. These plans cover our managed hosting, support, and hardened infrastructure — pick flat pricing or usage-based billing.
Starter
For a single API and a small team that just wants eyes on traffic.
- 1 monitored API
- Rule-based + statistical detection
- Email alerts
- 7-day log retention
- Community support
Pro
For teams running several APIs who need the full detection pipeline.
- Up to 10 monitored APIs
- Full detection pipeline incl. ML + local LLM
- Email, webhook & Slack alerts
- 90-day log retention
- Role-based access control
- Priority support
Pay As You Use
Billed on the requests you actually send through Boing — no seat count, no flat commitment. Ideal for spiky or seasonal traffic.
Final bill depends on your actual monthly event volume.
- Unlimited monitored APIs
- Billed per million ingested events
- Full detection pipeline
- Retention scales with usage
- Cancel or scale anytime
Questions, answered
Can't find what you're looking for? Open an issue on GitHub and we'll get back to you.