Lower pricing with Active CPU pricing for Fluid compute
Not a pilot. Vercel just cut LLM inference costs in half with Active CPU pricing.
Why it matters
Vercel's new Active CPU pricing model for Fluid Compute directly reduces operational spend for AI workloads by eliminating idle-time billing—critical economics for cost-conscious teams running LLM inference and long-running agents at scale.
The key facts
10 to knowActive CPU pricing reduces example workload cost from $0.31842/hour to $0.149/hour (~53% reduction)
Pricing structure: $0.128/hour Active CPU + $0.0106/GB-hour Provisioned Memory + per-invocation charges
Enabled by default for Hobby, Pro, and new Enterprise teams; existing Enterprise availability varies
Targets LLM inference, AI agents, and other workloads with significant idle time
Changes take effect after redeploy
Active CPU pricing reduces costs from $0.31842/hour to $0.149/hour for Standard machine at 100% utilization (53% savings)
Charges only for active CPU time ($0.128/hour), not idle periods
Provisioned memory billed separately at $0.0106 per GB-hour
Enabled by default for Hobby, Pro, and new Enterprise teams
Use case: LLM inference, long-running AI agents, workloads with significant idle time
Go to the source
Vercel Blogvercel.com
Publisher excerpt: Vercel Functions on Fluid Compute now use Active CPU pricing, which charges for CPU only while it is actively doing work. This eliminates costs during idle time and reduces spend for workloads like LLM inference, long-running AI agents, or any task with idle time. Active CPU pricing is built on…