Skip to content
CDN

Global network

Close to everyone, on six continents

988 edge locations, clustered around 65 metro hubs, put your content within 20ms of 95% of the world's internet users. Every deployment uses all of them.
Start freeBrowse regions
Edge locations
988
Continents
6
Median global latency
<15ms
Of users within 20ms
95%

Regions

Metro hubs by region

These are the metros where most capacity sits. The 988 locations are spread in and around them, run by independent operators, so traffic rarely has to leave the region it started in.

Latency shown is the typical round-trip to the nearest location in each region. See live measurements.

  • Europe

    18 metro hubs

    <10mstypical latency
    • Amsterdam
    • London
    • Frankfurt
    • Paris
    • Stockholm
    • Madrid
    • Milan
    • Warsaw
    • Dublin
    • Brussels
    • Vienna
    • Zurich
    • Copenhagen
    • Oslo
    • Helsinki
    • Prague
    • Bucharest
    • Athens

  • North America

    24 metro hubs

    <15mstypical latency
    • New York
    • Los Angeles
    • Chicago
    • Dallas
    • Miami
    • Seattle
    • Denver
    • Atlanta
    • Phoenix
    • Toronto
    • Montreal
    • Vancouver
    • San Francisco
    • Boston
    • Washington DC
    • Minneapolis
    • Houston
    • Detroit
    • Philadelphia
    • San Jose
    • Las Vegas
    • Salt Lake City
    • Mexico City
    • Guadalajara

  • Asia Pacific

    12 metro hubs

    <20mstypical latency
    • Singapore
    • Tokyo
    • Hong Kong
    • Sydney
    • Seoul
    • Mumbai
    • Jakarta
    • Bangkok
    • Manila
    • Taipei
    • Auckland
    • Melbourne

  • Latin America

    6 metro hubs

    <25mstypical latency
    • São Paulo
    • Buenos Aires
    • Santiago
    • Bogotá
    • Lima
    • Quito

  • Middle East

    3 metro hubs

    <20mstypical latency
    • Dubai
    • Tel Aviv
    • Bahrain

  • Africa

    2 metro hubs

    <30mstypical latency
    • Johannesburg
    • Cape Town

Nearest node

One address, answered by the closest location

Your hostname resolves to the same anycast address everywhere. The network delivers each visitor to the nearest healthy location, so a shopper in Kraków and one in Kyoto hit different nodes without you configuring anything.
  • No per-region setup, no geo-DNS rules to maintain
  • Unhealthy locations drop out automatically
  • Regional traffic split shows up in your analytics
Regional analytics docs

FAQ

Coverage questions

Do I choose which regions serve my content?
No. Every deployment is served from all locations, and each visitor is routed to the nearest healthy one. There's nothing to enable per region.
Why 988 locations but only 65 cities listed?
The list shows metro hubs. Locations are run by independent operators in and around those metros, so a single city can hold many of them.
Does bandwidth cost more in some regions?
No. Egress is free everywhere, and requests are priced the same in every region. Traffic served from São Paulo costs the same as traffic served from London.
Is my content stored in every location?
Cached content is stored where it's requested, so popular files end up everywhere quickly. To place files at every location up front, pin them with Edge Cache.

Deploy globally in two minutes

Create a deployment and your content is cached across 988 locations as visitors ask for it. No regions to pick, no per-region prices.

Start freeSee pricing

Free tier: 3 deployments and 500k requests a month, hard-capped. Bandwidth is always $0.