Railway vs Vercel pricing
Both of these are in Developer tool, and we read both pricing pages the same way, with the same reader. Everything below is what each page printed on the day we read it. Railway read on 13 Nov 2026, Vercel on 9 Nov 2026.
We check each page about once a week, so this is a weekly reading, not a live feed. Prices are copied as printed, never converted and never averaged.
The two pages
Vercel
Tier by tier
Rows are lined up by position on each page, not by tier name, because the names never match across two vendors. Row 1 is the top tier on each page, row 2 the one under it. Where one page has fewer tiers than the other, the row is left blank rather than paired with something it is not.
- Row 1RailwayFreeFreeVercelHobbyFree
- Row 2RailwayHobby$5/moVercelPro$20/mo
- Row 3RailwayPro$20/moVercelEnterpriseCustom
- Row 4RailwayEnterpriseCustomVercelNo tier at this position.
The entry tier
Railway is the cheaper of the two at the entry tier: $5/mo on Hobby, against $20/mo on Pro. That is $15 less, 75 percent below Vercel.
Both figures are for the whole account, per month, so they are like for like. This is the cheapest paid monthly tier we can read on each page. Free tiers, quoted tiers and per-unit rates are left out, because none of those is a figure you can put next to another company's figure.
Changes we recorded
Every change we have written up for either page, newest first.
Railway pricing page changed
Railway13 Nov 2026Vercel shifts from Edge Requests to CDN Requests model, introduces Flat Rate CDN, Vercel Drives, Daemons, and Backends; replaces per-GB overage billing with new pricing structures
Vercel8 Nov 2026Vercel shifts from Edge Requests to CDN Requests model, introduces Flat Rate CDN, Vercel Drives, Daemons, and Backends; replaces per-GB overage billing with new pricing structures Why it matters: • Vercel is fundamentally restructuring its billing model away from usage-based Edge overage charges ($2 per 1M, $0.15 per GB) to a Flat Rate CDN option with traffic spike protection, reducing customer unpredictability and competitive risk on variable workloads • New product categories (Vercel Drives, Daemons, Backends, Vercel CI) expand the platform into data persistence, background jobs, and compute-moving beyond edge deployment and raising switching costs for customers • Removal of Session Tracing and Advanced Deployment Protection ($150/month) and changes to Speed Insights (now free with 10K events/team, paid add-on for overages) signal consolidation and repositioning of observability offerings Pricing & Billing Model: - Edge Requests pricing model completely removed (was: 10M/month included, then $2 per 1M overage; 1TB/month included, then $0.15 per GB) - New Flat Rate CDN introduced as alternative to usage-based CDN billing on Pro plan, advertised as fixed monthly price with traffic spike protection and no overages - CDN Requests metric introduced (Included in Flat Rate CDN) - Tracing Spans pricing added: 1M span units/month included, then $0.50 per 1M span units (replaces prior Session Tracing with 1 hour/1 day of traces) - Trace Retention changed to 14 days of traces (was: 1 day of traces in one tier) - Speed Insights Plus Events pricing changed to $20 per 1M events (was: $20 per 1M for Edge Requests; now separate add-on) Plans & Features Removed: - Session Tracing feature removed entirely - Advanced Deployment Protection ($150/month) removed - Speed Insights Plus Events no longer built-in to plans New Products & Features Added: - Vercel Drives: Drive Storage 15 GB lifetime included, then $0.05 per GB-month; Drive Reads 30 GB/month included, then $0.0015 per GB; Drive Writes starting at $0.004 per GB - Vercel Daemons: Active CPU and Provisioned Memory tiers added - Backends product category introduced - Vercel CI: 200 CPU minutes/month included - Deployment Storage: 10 GB included - Marketplace Integrations added - Custom Monitors: $0.0035 per GB scanned - Password Protection: $20/project/month Observability & Monitoring: - Speed Insights repositioned as Free tier with 10K events/team/rolling 30 days - Speed Insights Plus Events now available as paid add-on (pricing structure changed) - Tracing Spans introduced as managed feature with 1M span units/month included, $0.50 per 1M overage Positioning & Messaging: - FAQ section added: 'What is Flat Rate CDN?' with explanation of fixed pricing, spike coverage, and no-overages value prop - Copy shift from 'Edge Requests' terminology to 'CDN Requests' throughout - Detailed caching/compression explanation retained but updated to reference CDN Requests instead of Edge Requests What to do: • Audit your own CDN and overage billing model: if you charge per-request or per-GB with variable overages, Vercel's Flat Rate CDN is now a direct competitive threat. Evaluate offering a predictable tier or capping overage costs • Assess whether your observability stack (Speed Insights, tracing, session replay) can compete with Vercel's expanded free Speed Insights + paid add-on model; determine if you need feature parity or differentiation in monitoring/debugging • Map Vercel's new products (Drives, Daemons, Backends, CI) against your platform capabilities-these are land-and-expand vectors. Prioritize blocking the most dangerous combination (e.g., Drives + Daemons for stateful workloads) or lean into areas you own
Railway introduces Sandboxes & VMs as a new product offering with per-second metering at $50/GB-month and $50/vCPU-month, removes Cloud Regions from Enterprise tier.
Railway8 Nov 2026Railway introduces Sandboxes & VMs as a new product offering with per-second metering at $50/GB-month and $50/vCPU-month, removes Cloud Regions from Enterprise tier. Why it matters: • Railway is expanding beyond traditional container deployment into infrastructure primitives (VMs/sandboxes), signaling a shift toward broader compute workloads and agent/code execution use cases that could capture workloads currently served by competitors like AWS Lambda or specialized VM providers. • The removal of Cloud Regions from Enterprise tier (now listed only as 'Higher Resource Limits') suggests Railway is either consolidating features or pricing strategy; competitors should clarify their own cloud region strategy and pricing transparency. • The sub-second metering model ($0.00001929 per GB/sec, $0.00001929 per vCPU/sec) and idle-only billing positioning is a direct cost-efficiency message; Railway is emphasizing pay-for-actual-use over wall-clock time, which could appeal to cost-conscious workloads. Product/Features: - New Sandboxes & VMs offering launched: isolated Linux VMs for agents, builds, and untrusted code execution. - Sandbox billing model: metered per second at $0.00001929 per GB/sec and $0.00001929 per vCPU/sec, approximately $50 per GB-month and $50 per vCPU-month for active use only. - Sandbox lifecycle features added: checkpoints, forks, templates, private networking, and public domains. - Idle timeout feature: configurable up to 5 minutes (default), up to 2 hours, or never, with automatic destruction of idle sandboxes. - Sandbox resource tiers: default 2 vCPU/2 GB, scaling to maximum 32 vCPU/32 GB per sandbox. Pricing: - Sandbox pricing: $50 per GB of memory per month and $50 per vCPU per month, metered per second, plus $0.05 per GB egress. - Pricing model emphasizes actual usage: customers pay only for CPU actually used and memory held while running, not for idle wall-clock time. Plans: - Enterprise tier: Cloud Regions removed from listed benefits; now shows 'Enterprise Support, Higher Resource Limits' plus prioritized support with SLOs, dedicated Slack Connect channel, and services with 48vCPU/48GB RAM per replica. Positioning: - Railway is positioning sandboxes as a cost-efficient alternative for waiting/idle workloads ('a sandbox waiting on a model or a user costs almost nothing'). - Emphasis on flexible, per-second metering and automatic cleanup (destroyed sandboxes cost nothing) versus traditional VM hourly billing. What to do: • Audit your own VM/sandbox offering (if any) and cost messaging: clarify whether you meter per-second or per-hour, and whether you charge for idle time. Railway's positioning on 'pay-only-for-active-use' is now a competitive liability if you don't match it. • Evaluate whether Railway's removal of Cloud Regions from Enterprise signals a pricing restructuring or de-emphasis; confirm your own multi-region story with customers and sales to prevent churn justifications. • Monitor Railway's Sandboxes adoption and feature velocity; if this becomes a revenue driver, assess whether your product roadmap should include similar compute-primitive offerings or deeper integrations with agent/LLM workflows that depend on untrusted code execution.
Vercel introduces per-minute pricing for Basic machine compute, starting at $0.007/min, expanding infrastructure monetization
Vercel3 Sept 2026Vercel introduces per-minute pricing for Basic machine compute, starting at $0.007/min, expanding infrastructure monetization Why it matters: • Vercel is now monetizing granular compute resources that were previously bundled into plans, establishing a new revenue stream and potentially raising TCO for customers running baseline workloads • This represents a shift toward infrastructure-as-you-go pricing beyond traditional seat/plan models, which could impact competitive positioning if customers perceive hidden costs or usage uncertainty • The entry point of $0.007/min creates new pricing complexity that competitors should clarify in their own pricing communication to avoid customer confusion or dissatisfaction Pricing: - New compute billing introduced: Basic machines at $0.007 per minute (exact prior pricing unknown from diff, but this is a new line item) Plans: - Basic machines now explicitly listed as 'Included' in a plan tier, suggesting tiered access or quota-based compute allocation Positioning: - Shift toward consumption-based pricing for infrastructure resources, moving beyond flat-rate plan tiers toward granular pay-as-you-go for compute What to do: • Audit own pricing strategy: determine whether bundled compute limits or transparent per-minute billing is clearer and more competitive for core customer segments • Prepare customer comms addressing compute costs and limits-if Vercel's per-minute model creates sticker shock, emphasize predictability of your own pricing model • Monitor Vercel's adoption metrics and customer churn post-launch to validate whether consumption-based pricing drives upsell or customer friction
Railway adds explicit usage-based pricing transparency and enterprise demo booking to pricing page
Railway3 Sept 2026Railway adds explicit usage-based pricing transparency and enterprise demo booking to pricing page Why it matters: • Railway is emphasizing pay-for-what-you-use cost predictability as a competitive differentiator, signaling confidence in their resource efficiency to undercut traditional flat-rate competitors • The addition of enterprise demo booking directly on the pricing page suggests they're optimizing conversion for high-value deals and moving away from pure self-serve positioning • New messaging around '0% idle cost' and 'only pay for active CPU/memory' indicates Railway is repositioning against competitors with wasteful resource billing or hidden charges Positioning: - Added explicit messaging: 'Only pay for active CPU and memory' (suggests focus on resource efficiency vs. competitors) - Added '0% idle cost' messaging (new value prop differentiator) Pricing: - Introduced or emphasized usage-based billing transparency with specific metrics (CPU/memory-based rather than generic 'compute credits') What to do: • Audit your own pricing model messaging-if usage-based, ensure it matches or exceeds Railway's transparency on what 'active' means and how costs scale • Review enterprise sales funnel: Railway's demo booking link suggests they're converting high-touch deals; ensure your enterprise onboarding process is equally frictionless • Test messaging that counters 'pay for idle time'-either highlight your own utilization efficiency or introduce usage-based tiers if you haven't already
Vercel pricing page changed
Vercel2 Sept 2026Railway pricing page changed
Railway1 Sept 2026Vercel pricing page changed
Vercel27 Aug 2026Vercel pricing page changed
Vercel25 Aug 2026Vercel pricing page changed
Vercel22 Aug 2026Vercel rebrands Edge Config to Global Config across all pricing tiers
Vercel17 Aug 2026Vercel rebrands Edge Config to Global Config across all pricing tiers Why it matters: • Positioning shift suggests Vercel is broadening the perceived scope of this feature beyond edge-specific use cases to appeal to broader infrastructure needs • Renaming may indicate feature maturation and expansion-competitors should audit whether their own edge/global config offerings are sufficiently differentiated • The terminology change ('Global' vs 'Edge') could influence buyer perception of feature capabilities and use-case applicability in competitive evaluations Features: - Edge Config renamed to Global Config - Edge Config Reads renamed to Global Config Reads - Edge Config Writes renamed to Global Config Writes What to do: • Audit competitive positioning of your own configuration management features-ensure messaging clarity around edge vs. global scope to differentiate from Vercel's repositioning • Monitor whether Vercel expands Global Config capabilities post-rebrand (read/write limits, latency SLAs, integrations) that could affect feature parity claims • Update internal battlecards and sales materials to address the rebrand in customer conversations to prevent confusion during competitive deals
Watch both of these yourself
Tierpeek runs this same read on any pricing page you paste below, every 8 hours, and sends what moved to your inbox or Slack.
Free for 3 days, no card. Or create an account first and add pages from the dashboard.