VMware License Costs 2026: What SMBs Actually Pay

Since Broadcom acquired VMware in November 2023, the licensing model has been completely rebuilt: away from perpetual licences, towards subscription bundles with minimum commitments. The first renewal letters hit customers in early 2024, and the topic has been present in every second SMB advisory since. Almost two and a half years later the situation is stable — and not in favour of smaller customers.

This article deliberately avoids made-up numbers. Instead: what we actually see in DATAZONE consulting, how the current licensing model is structured, and which alternatives are realistic for SMBs.

The 2026 licensing model — structural overview

Broadcom has compressed the previous portfolio (ESXi, vCenter, vSAN, NSX, Tanzu as separately licensable products) into two subscription bundles:

Bundle	Contents	Target
VMware vSphere Foundation (VVF)	vSphere, vCenter, Aria Operations, Tanzu Kubernetes base	SMB, classic virtualisation
VMware Cloud Foundation (VCF)	VVF + vSAN + NSX + full Aria stack	Enterprise, private cloud stack

Structural changes that hit SMBs hardest:

• Mandatory subscription: perpetual licences are no longer sold. Existing customers with perpetual rights lose access to updates and security patches when their support contract ends.
• Minimum-core commitment: currently 16 cores per CPU as the floor — even if the actual CPU has fewer cores. Small hosts with 8-core Xeons pay a meaningful uplift.
• Bundle obligation: individual products (e.g. vSphere Standard on its own) are no longer in the main programme. Anyone who only wanted hypervisor + vCenter now buys VVF — including Aria components they may not need.
• Channel reduction: many smaller resellers have lost or voluntarily handed back their VMware authorisation. Procurement runs through significantly fewer providers.

For exact 2026 list prices, ask an authorised Broadcom partner directly — we deliberately do not quote numbers here because terms differ by customer and region, and list prices are not identical to end-customer prices.

What we actually see in consulting

At DATAZONE we typically advise SMB IT leads running between 2 and 20 ESXi hosts. The recurring patterns:

Pattern 1: renewal shock after a 3-year contract Customers whose last multi-year contract expired in 2024/2025 see a three- to four-digit percentage increase at renewal over the old contract. The most common consulting question: “is that a negotiation opening or a real demand?” — answer: mostly real, with limited negotiation room for multi-year commitments.

Pattern 2: mini-cluster effect 3-node ESXi clusters with small CPUs (8–12 cores per socket) are heavily affected by the 16-core minimum. They pay for cores they physically do not have.

Pattern 3: loss of reseller trust Where the existing IT service provider has lost Broadcom authorisation, a partner change is added on top. That can double the migration effort: new platform and new partner at the same time.

Pattern 4: forced strategy decision Customers who treated VMware as “it just runs” for ten years suddenly need an actual platform strategy. That is exhausting in the short term, but healthy long term.

Realistic alternatives for SMBs

Three paths are actually viable in SMB practice — each with pros and cons:

Alternative 1: migration to Proxmox VE

Proxmox VE is the obvious open-source alternative and has been production-ready for years. The DATAZONE migration practice since 2024 shows: for classic SMB virtualisation, Proxmox 9.x covers roughly 95% of vSphere Standard’s functionality. Where gaps remain, they are usually features that SMBs do not actively use (complex Distributed Switch in NSX setups, EVC cluster migrations, vRealize tooling).

Strengths:

• No licence subscription, optional support contract
• Active community, vendor based in Austria
• Full ZFS, Ceph and SDN features without add-on modules
• Native TrueNAS integration since early 2026

Weaknesses:

• Real migration effort per VM (disk conversion, config transfer)
• HA cluster setup needs knowledge not every IT team brings along
• No 1:1 DRS replacement — the Dynamic Load Balancer in PVE 9.2 is solid but younger than DRS

When it makes sense: for 3 to ~30 hosts with classic VM workloads (Windows Server, databases, Linux). DATAZONE typically plans migration in 4–12 weeks depending on cluster size and complexity.

Alternative 2: move to Hyper-V

If your workload is mostly Windows Server and you already deploy Microsoft licences broadly, Hyper-V can be a sensible platform check. Cost-wise Hyper-V is not “free” — Windows Server Datacenter per host is non-trivial, but then covers live migration, failover clustering and all current Hyper-V features.

Strengths:

• Clean Microsoft stack integration (Active Directory, System Center, Azure hybrid)
• Familiar management tools, low learning curve for Windows admins
• Stable live migration and failover cluster mechanisms

Weaknesses:

• Linux workloads run, but it is not their preferred habitat
• Storage stack (S2D, CSV) more complex than Proxmox ZFS/Ceph
• Microsoft’s own licensing trajectory is not static — see the ongoing Microsoft 365 discussion

When it makes sense: when the infrastructure is already Microsoft-centric and there are strong organisational reasons against open source.

Alternative 3: stay on VMware — deliberately

Staying on VMware is a legitimate option if:

• The workload is business-critical and tied to specific VMware features (e.g. SRM replication, NSX microsegmentation, Aria Operations compliance reports)
• The renewal negotiation lands at terms comparable to migration TCO
• Internal IT has no capacity for a platform migration in the next 12 months

Important: “staying on VMware” should be a conscious, documented decision — not “we’ll wait and look again in 2027”. The next pricing round is coming.

Decision matrix

Pragmatic guidance by host count and strategy:

Situation	Recommendation
1–2 hosts, simple VM load	Proxmox — migration in 2–4 weeks, ROI by first renewal
3–10 hosts, classic SMB	Proxmox, migration planned 6–12 weeks
5–15 hosts, heavily Windows-centric	Proxmox or Hyper-V depending on staffing
10+ hosts with active NSX/vSAN/SRM	Negotiate with Broadcom first, evaluate pilot migration in parallel
Compliance workload with VMware-specific certification	Stay on VMware, renegotiate, prepare exit strategy quietly
2–4 hosts running “somehow” on perpetual	Plan now — security patches end with support contract

What an honest TCO calculation must include

A “Proxmox saves X EUR” calculation without migration cost is not honest. Realistic 3-year TCO comparisons should factor in:

1. Migration effort: consulting + host conversion + VM migration + testing + IT team training
2. Storage strategy: VMware VSAN drop-out means either Proxmox Ceph, traditional SAN, or TrueNAS as external storage
3. Backup adaptation: existing Veeam/PBS strategy adapted to the new platform
4. Training and doc updates: runbooks, DR plans, monitoring templates
5. Support contract on the new platform: Proxmox enterprise subscription, Hyper-V Premier or continued VMware subscription

Experience from DATAZONE migrations: migration costs amortise for SMB VMware customers usually within the first subscription period — i.e. 3 years. For small setups often faster.

What Broadcom negotiation realistically achieves

In consulting we see consistently: negotiating is worth it — but only inside a defined corridor. Concretely:

• Multi-year commitments (3 or 5 years) typically bring noticeable discounts vs. annual subscription
• Larger volumes (voluntary higher core counts) bring percentage discounts
• Credible alternative on the table (e.g. documented Proxmox pilot) measurably improves negotiating position
• Pricing close to old perpetual terms is not achievable — that is strategically by design

Anyone negotiating seriously should define a walk-away position — i.e. know in advance from which renewal price point migration becomes realistically cheaper.

DATAZONE recommendation

For the majority of our SMB customers: plan the Proxmox migration now, execute over the next 6–18 months. We have run many of these migrations since 2024 — the pattern works, the platform is stable, and the switch often resolves latent storage topics alongside the licensing one (TrueNAS integration, 3-2-1 backup, clean snapshot strategy).

But: if you actually need a specific VMware feature and the renewal lands within an acceptable range after negotiation, staying is legitimate. We advise vendor-neutrally — even when that argues against a migration project for us.

Want a concrete assessment? We provide a position statement for your environment: current licence state, planned workloads, honest comparison of the three options above — as the basis for board-level decision. More under initial consultation.

Related articles

• From VMware to Proxmox — DATAZONE migration guide
• Proxmox vs. VMware — the honest comparison
• VMware and Broadcom: background to the acquisition
• Proxmox VE 9.2 — current features
• TrueNAS + Proxmox: official plugin