All systems operational

Edge infrastructure
for latency-sensitive APIs.

Dipolle routes API traffic through a private network of edge nodes, serving responses from the closest healthy region in under 50 ms. No new clients. No config drift. No surprises.

99.99%
Uptime, last 12 months
42ms
Median edge response
18
Regions worldwide
2.4B
Requests served / day
Platform

Built for teams that can't afford a bad second.

Three primitives compose the entire platform. Everything else is configuration, not architecture.

Edge routing

Anycast-based entry points in 18 regions. Traffic arrives at the nearest PoP and reaches its destination through a private backbone — not the public internet.

Real-time analytics

Every request streamed into your telemetry pipeline in under a second. Structured, searchable, replayable. No sampling by default.

Private by default

mTLS on every hop. Per-environment keys. SOC 2 Type II. Zero-trust doesn't need to be a checklist — just an assumption.

Automatic failover

Health-aware routing at the protocol layer. A regional outage becomes a 200 ms blip, not a page.

Drop-in integration

Point your DNS at Dipolle. That's it. No SDK, no agent, no sidecar. Our proxies speak your protocols.

Predictable pricing

Bandwidth-based, not per-request. A hockey-stick in growth shouldn't mean a hockey-stick in invoices.

Developer experience

Ten lines of config. Zero surprises.

Declare your service, regions, and health checks. Dipolle handles the rest — propagation, failover, observability.

View documentation
dipolle.yml
service: api.acme.io
regions: [iad, fra, sin]

routes:
  - match: /v1/*
    upstream: origin.acme.internal
    timeout_ms: 800
    retry: 2

health:
  path: /_healthz
  interval: 5s
  grace: 30s

# Applied in ~200 ms globally.