AWS Lambda Pricing Plans and Tiers

AWS Lambda's pricing model ($0 free tier with 1M monthly requests, per-execution billing thereafter) represents the paradigm of consumption-based serverless pricing. The free tier removes evaluation friction, while the per-execution model ($0.20 per million requests, $0.00001667 per 100ms execution time) creates cost unpredictability that frustrates startups. A viral application incurring 1 billion monthly requests would cost $200K in request fees plus execution charges, making Lambda budgeting an exercise in worst-case scenarios.

AWS Lambda's strength is integration breadth: direct event sources (S3, DynamoDB, API Gateway, SNS, SQS) enable serverless workflows without external orchestration. However, this tight coupling increases switching costs and platform lock-in. Teams that build Lambda-centric architectures across multiple AWS services find migration economics prohibitive, enabling AWS to extract premium pricing in year 2-3 of production deployments.

Lambda's per-execution billing incentivizes optimization but creates operational complexity: teams must constantly tune memory allocation, timeout settings, and cold start behavior to minimize cost. This optimization burden is an invisible operational cost that competes against flat-rate platforms (Cloudflare Workers $5/mo, Deno Deploy $20/mo) that trade margin efficiency for simplicity.