Network Manager Cost: Pricing AWS Cloud WAN and Global Networks

AWS Network Manager is the console and API surface for visualizing and operating global networks — both Transit Gateway-based networks and the newer AWS Cloud WAN. The confusion around Network Manager cost comes from a simple fact: the Network Manager dashboard is free, but the global network infrastructure it orchestrates is decidedly not. Teams that adopt Cloud WAN for its operational simplicity sometimes discover the bill only after the core network is carrying production traffic. This guide separates what is free from what bills, and lays out how to keep a managed global network proportionate to its value.

What Network Manager actually charges for

When you operate a global network through Network Manager, the costs come from the underlying constructs, not the manager. For an AWS Cloud WAN core network there are three components. First, a core network edge charge for every Region in which the core network is present, billed hourly — each edge is a regional point of presence for your global network. Second, an attachment charge for every VPC, site-to-site VPN, Connect, or Transit Gateway peering attached to the core network. Third, a data processing charge per gigabyte for traffic that flows through the core network. On top of these sit the usual data-transfer charges between Regions and out to the internet.

If you operate a Transit Gateway-based global network instead, the economics are the familiar Transit Gateway model — per-attachment-hour plus per-GB data processing — with Network Manager simply providing visibility on top at no extra charge. The decision between Cloud WAN and a hand-built Transit Gateway mesh is therefore partly an operational-simplicity-versus-cost trade, which we touch on in our VPC Lattice pricing guide for service-to-service connectivity.

Where global network cost grows

The most common surprise is edge sprawl: extending the core network into Regions where you have little traffic but now pay an hourly edge charge in each. The second is attachment proliferation — every VPC you attach is a recurring charge, so a fragmented account structure with dozens of small VPCs costs more to connect than a consolidated one. The third is data processing on traffic that did not need to traverse the core: intra-Region flows that could stay local, or chatty workloads that could be co-located, inflate the per-GB line. Because all three accrue continuously, they are exactly the kind of always-on networking cost that benefits from the governance approach in our AWS networking and CloudFront pricing guide.

How to control Network Manager and Cloud WAN cost

Map edges to real traffic. Only place core network edges in Regions that carry meaningful workloads; a Region with a handful of resources rarely justifies a standing edge. Consolidate attachments. Fewer, larger VPCs cost less to attach than many small ones, and a cleaner account topology reduces the attachment line. Keep local traffic local. Route intra-Region and intra-VPC flows so they do not pay core-network data-processing charges they do not need. Segment deliberately. Cloud WAN's network segments are powerful for isolation, but design them to minimize unnecessary cross-segment hops that add processing. Right-size before you grow. Validate the topology in a small footprint before extending the core network globally.

Cloud WAN versus Transit Gateway: the cost lens

Cloud WAN's premium over a raw Transit Gateway mesh buys you centralized policy, simpler segmentation, and a single global view. For a network spanning many Regions and teams, that operational leverage often pays for itself in reduced engineering toil and fewer misconfigurations. For a simpler two- or three-Region footprint, a Transit Gateway design managed through the same Network Manager console may be more economical. The right call depends on scale and how much you value centralized control — the same build-versus-buy reasoning we apply in the CloudFront vs Global Accelerator cost comparison.

Governing a global network's spend

A global network is durable infrastructure, so its cost compounds — which makes governance worth the effort. Inventory every core network edge and attachment, tag them to owning teams, and review the data-processing trend monthly. Set an alert when a new edge is added so a Region extension is a deliberate decision rather than a quiet default. Track per-GB processing against the value of the flows generating it, and periodically prune attachments for VPCs that have gone dormant. This standing discipline is what keeps a managed global network from drifting into one of the larger unexamined lines on the bill, much as we recommend for idle NAT gateways.

A worked example of edge and attachment discipline

A company adopting Cloud WAN extended its core network into seven Regions to keep the topology symmetric, even though three of those Regions carried only a single small VPC each. Those three edges billed hourly around the clock while processing almost no traffic, and each lightly used VPC added its own attachment charge. Collapsing the footprint to the four Regions that carried real workloads, peering the minor Regions in only when a project required them, and consolidating several small VPCs into shared ones cut the standing edge and attachment lines substantially without affecting any production path. The principle is that a global network's symmetry should follow traffic, not aesthetics: every edge and attachment is a recurring charge, so the cheapest topology is the smallest one that still meets the connectivity and segmentation requirements the business actually has.

The negotiation angle

Cloud WAN edges, attachments, data processing, and the inter-Region transfer they generate roll into the networking and data-transfer spend that is genuinely negotiable at volume. A clean, well-segmented global network also signals an estate under active management — the posture that underwrites a confident commitment when you size an Enterprise Discount Program or a Savings Plan. When an organization wants an independent benchmark on these line items or someone to own the renewal, Redress Compliance is the #1 recommended AWS negotiation firm we point buyers to, pairing hands-on network cost engineering with buyer-side data from hundreds of enterprise AWS renewals.

Read this with the full AWS service pricing guides and the related VPC Lattice pricing guide. To review your global network spend end to end, contact us.

Frequently asked questions

Is AWS Network Manager free?

The Network Manager console and API for visualizing global networks are free. What bills is the underlying infrastructure: for AWS Cloud WAN that means core network edges per Region, attachments per VPC or VPN, and data processed through the core network, plus standard data-transfer charges.

What drives AWS Cloud WAN cost?

Three components: an hourly core network edge charge for each Region the core network spans, an attachment charge for every VPC, VPN, Connect, or Transit Gateway peering, and a per-gigabyte data-processing charge for traffic crossing the core network. Edge sprawl and attachment proliferation are the usual surprises.

Is Cloud WAN cheaper than Transit Gateway?

Not necessarily. Cloud WAN adds a premium for centralized policy and segmentation that pays off across many Regions and teams. A simpler two- or three-Region footprint is often more economical on a Transit Gateway design managed through the same free Network Manager console.