Introduction to this benchmark
This analysis evaluates the cost-efficiency of 7 edge serverless providers by simulating three distinct traffic profiles: Entry, Growth, and Scale. By measuring Total Cost of Ownership (TCO) against specific request and bandwidth parameters, we isolate the financial impact of architectural choices. For engineering leaders, tracking TCO is essential to prevent "serverless sprawl," where hidden egress and storage overages negate the operational savings of managed platforms.
Executive summary
The current edge serverless market, represented by the 7 providers analyzed, exhibits a bifurcated pricing structure: low-barrier entry tiers versus aggressive consumption-based overage models. While base pricing often appears negligible, the primary cost driver for scaling workloads is network egress and storage throughput, which can cause TCO to deviate significantly from advertised monthly rates.
Engineering teams must distinguish between unmanaged IaaS, which requires higher DevOps investment but offers granular control, and fully-managed platforms that trade higher unit costs for reduced operational overhead. Our analysis reveals that providers with zero-egress fees, such as Cloudflare , offer superior predictability for high-bandwidth applications compared to competitors with high-margin egress penalties.
Methodology and workload definitions
- Entry/Hobby: 1 request and 10 units of bandwidth.
- Growth/Startup: 10 requests and 100 units of bandwidth.
- Enterprise/Scale: 50 requests and 1000 units of bandwidth.
📈 The TCO Scaling Curve: Real-time cost explosion curves and interactive workload simulations are available in the full interactive benchmark report on CloudCompare.
Scenario 1: Entry-level economics
At the Entry tier, the market is dominated by "freemium" models where Azure , Netlify , Cloudflare , Fly-io , GCP , and Vercel all offer a $0 TCO. This allows for rapid prototyping without immediate capital expenditure. However, the underlying cost structure varies significantly; Cloudflare stands out by eliminating egress and RAM costs entirely, whereas Fly-io introduces a $0.15 per unit storage fee and $0.02 per unit egress fee, which remain dormant at this scale but signal future liability.
Supabase represents the outlier in this tier with a $25 base price for its PRO plan. While this appears expensive compared to the $0 entry points of its peers, it provides a managed database-as-a-service layer that justifies the premium for teams lacking the DevOps capacity to manage their own persistence layer. Netlify and Vercel offer zero-cost entry but enforce strict usage limits that necessitate an immediate upgrade to paid tiers as soon as traffic patterns stabilize.
| Provider | Optimal plan | Base price | Entry TCO |
|---|---|---|---|
| Microsoft Azure | Azure Functions Serverless Plan | $0.00/mo | $0.00/mo |
| Netlify | Netlify Starter Free | $0.00/mo | $0.00/mo |
| Cloudflare | Cloudflare Workers Free | $0.00/mo | $0.00/mo |
| Fly.io | Fly Anycast Edge Serverless | $0.00/mo | $0.00/mo |
| Google Cloud (GCP) | GCP Cloud Functions Plan | $0.00/mo | $0.00/mo |
| Vercel | Hobby | $0.00/mo | $0.00/mo |
| Supabase | PRO | $25.00/mo | $25.00/mo |
Scenario 2: Growth stage scaling
As workloads transition to the Growth tier, the financial landscape shifts toward fixed-fee subscriptions. Netlify moves to a $19 base price, while Vercel shifts to $20. These providers prioritize developer experience and CI/CD integration, effectively charging a premium for the "zero-ops" environment. Cloudflare remains the most aggressive competitor here, maintaining a low $5 base price for its paid Workers tier.
The pricing models begin to diverge based on resource consumption. While Fly-io maintains a $0 base price even at the Growth tier, its reliance on consumption-based overages means that high-bandwidth applications will eventually exceed the cost of the fixed-fee plans offered by Netlify or Vercel. Engineering managers must decide if they prefer the predictability of a $20 monthly bill or the potential volatility of a consumption-only model.
| Provider | Optimal plan | Base price | Growth TCO |
|---|---|---|---|
| Fly.io | Fly Anycast Edge Serverless | $0.00/mo | $0.00/mo |
| Cloudflare | Cloudflare Workers Paid | $5.00/mo | $5.00/mo |
| Netlify | Netlify Pro Plan | $19.00/mo | $19.00/mo |
| Vercel | Pro | $20.00/mo | $20.00/mo |
Scenario 3: Enterprise scale and the overage trap
At the Enterprise Scale tier, the "overage trap" becomes the primary determinant of TCO. Netlify reaches a $99 base price, but its persistent overage fees of $0.55 per unit of egress represent a significant financial risk for high-traffic applications. This creates a scenario where a single traffic spike can result in costs that dwarf the base subscription fee, making Netlify a high-risk choice for bandwidth-heavy workloads.
Fly-io demonstrates a more balanced approach at scale, with a $18 TCO and lower overage penalties of $0.15 per unit of storage and $0.02 per unit of egress. This structure is significantly more sustainable for high-scale operations than the $0.15 per unit egress fee found on Vercel's Pro plan. The financial anatomy of these costs shows that providers with high egress penalties are effectively taxing the success of the application, whereas those with lower unit costs allow for linear scaling.
Ultimately, the most dangerous trap is the combination of a low base price and high unit overages. Providers like Netlify and Vercel are optimized for low-to-medium traffic, but their cost curves steepen sharply at scale. In contrast, Cloudflare's model—which eliminates egress and RAM costs—remains the most architecturally sound choice for enterprise-scale edge deployments, as it decouples cost from bandwidth consumption, providing the financial predictability required for large-scale infrastructure planning.
| Provider | Optimal plan | Base price | Scale TCO |
|---|---|---|---|
| Fly.io | Fly Anycast Edge Serverless | $0.00/mo | $18.00/mo |
| Netlify | Netlify Enterprise Benchmark | $99.00/mo | $99.00/mo |
Architectural verdict
For teams with limited DevOps resources, the premium fixed-fee models of Vercel and Netlify are recommended for their operational simplicity, provided traffic remains within predictable bounds. However, for high-scale, bandwidth-intensive applications, the consumption-based overage traps inherent in these platforms are prohibitive. We advise CTOs to migrate high-scale workloads to providers like Cloudflare , where the absence of egress fees provides a predictable, linear cost model that scales without the risk of exponential overage penalties.
🔗 Compare live infrastructure costs and run custom workload simulations: Open Interactive Cloud Calculator
Stop estimating. Start calculating.
Apply this benchmark data to your exact architecture. Use our deterministic multi-cloud calculator to forecast your database, PaaS, and egress costs with zero vendor bias.
Top comments (0)