AI BriefingAWS Lambda Web Functions: Billing Based on Active CPU Time and Support for 128 Concurrent Requests
·2026.10.29 11:46
Key point
A cost reduction of approximately 66% compared to existing solutions is confirmed for LLM streaming processing.
Details
- AWS Lambda Web Functions deploys standard HTTP servers without adapters and handles up to 128 concurrent requests per execution environment.
- The billing model charges only for active CPU time, and I/O wait time is not reflected in the cost.
- For LLM streaming workloads, monthly costs are reduced by approximately 66% compared to existing Event Functions.
- Only Node.js 24.x (Arm64) is supported, and the execution environment specifications are fixed at 2 vCPU and 2 GB of memory.
- Amazon CloudFront supports Web Functions as an origin and uses the dedicated signing action always-amz-auth.
- Available regions are US East/US West, Tokyo, and Ireland, and domestic users are recommended to use a combination of the Tokyo home region and the Seoul region endpoint.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
AWS Launches Lambda Web Functions for Standard HTTP ApplicationsAWS Lambda Web Functions enables standard HTTP servers without adapters; S3 available on Vercel Marketplace
Kimi K3.1 Pro Scores 55 on Artificial Analysis Intelligence Index