AWS Lambda cost varies with how often functions run, how long they execute, and the memory assigned. The main drivers are request count, compute duration, and data transfer. This guide provides practical pricing ranges in USD to help plan budgets and compare with alternatives.
| Item | Low | Average | High | Notes |
|---|---|---|---|---|
| Requests | $0.00 | $0.20 | $0.60 | First 1M requests per month are free; beyond that the price scales modestly. |
| Duration (Compute Time) | $0.00 | $0.50 | $4.00 | Charged per 100ms; depends on memory allocation (128MB–10GB). |
| Memory Allocation Increment | — | — | — | Higher memory increases cost per 100ms but may reduce duration. |
| Data Transfer (Out to Internet) | $0.00 | $0.10 | $0.50 | Charges apply after free data transfer allowances. |
| Provisioned Concurrency | $0.00 | $0.10 | $2.00 | Keeps functions warm; billed by provisioned amount and duration. |
Typical Cost Range
The typical monthly cost for a modest Lambda workload ranges from a few dollars to a few hundred dollars, depending on usage patterns and memory choices. For low-traffic apps with infrequent events, Lambda can be near-free within the free tier. For API-heavy or data-processing tasks, monthly bills commonly fall in the low–average–high bands below, assuming consistent traffic and ordinary data transfer.
Cost Breakdown
Key components are requests, compute time, and data transfer. The cost breakdown below shows a practical set of columns to estimate a Lambda bill. Assumptions: region is US East (N. Virginia); average function runs are short, with mixed memory settings; modest outbound data transfers.
| Item | Low | Average | High | Notes |
|---|---|---|---|---|
| Requests | $0.00 | $0.20 | $0.60 | First 1M free; cost scales with monthly demand. |
| Duration | $0.00 | $0.50 | $4.00 | Billing per 100ms; higher memory increases per-unit rate. |
| Data Transfer Out | $0.00 | $0.10 | $0.50 | Standard egress charges apply beyond free allowances. |
| Provisioned Concurrency | $0.00 | $0.10 | $2.00 | Costs scale with configured concurrency and duration. |
| Monitoring & Logs | $0.00 | $0.05 | $0.50 | CloudWatch logs and metrics incur minor fees. |
Factors That Affect Price
Usage pattern and memory allocation are the biggest price levers. The two niche drivers to monitor are event volume (invocation count) and function memory settings (128MB–10GB). A higher memory configuration increases cost per unit of compute time but can reduce total duration, potentially lowering overall spend for CPU-bound tasks. Additionally, outbound data transfer and any optional features like Provisioned Concurrency or dedicated SLAs influence the total.
Regional Price Differences
Prices vary modestly by region. In the United States, the most common comparisons are US East (N. Virginia) versus US West (Oregon) and US East (Ohio). Typical deltas from the U.S. Northeast baseline are within ±5% to ±15% depending on data transfer routes and regional pricing for services that accompany Lambda usage (e.g., API Gateway, data egress). Enterprises with global endpoints may see regional variance in chargeable bandwidth and cold-start considerations.
Real-World Pricing Examples
Three scenario cards illustrate daily ranges based on common workloads.
-
Basic Webhook Bot — 500,000 requests/month, average 120ms duration at 256MB memory, 10MB outbound data per request.
Assumptions: region US East, no provisioned concurrency.- Invocations: 0.5M
- Avg duration: 120ms
- Memory: 256MB
- Data out: 5MB
- Total estimate: $6–$12 per month
-
API Backend — 3M requests/month, average 250ms, 512MB memory, 100MB data out monthly, modest logs.
Assumptions: provisioned concurrency off.- Invocations: 3M
- Avg duration: 250ms
- Memory: 512MB
- Data out: 100MB
- Monitoring: included
- Total estimate: $60–$150 per month
-
Data Processing Job — 2M events, 1.5s duration, 1GB memory, 2GB data out monthly; optional Provisioned Concurrency.
Assumptions: region US East, some cold starts avoided with provisioning.- Invocations: 2M
- Avg duration: 1.5s
- Memory: 1024MB
- Data out: 2GB
- Provisioned Concurrency: enabled for peak hours
- Total estimate: $200–$900 per month
What Drives Price
Compute efficiency and event-driven design drive the value of Lambda. Architectures that minimize idle time, favor shorter durations, and reduce data transfer costs tend to be cost-effective. When workloads experience spikes, evaluate Provisioned Concurrency or consider alternative architectures for steady heavy usage. The decision hinges on the balance between predictable traffic, latency requirements, and the cost of always-on capacity versus on-demand execution.
Ways To Save
Several practical tactics can lower AWS Lambda spend without compromising value. Leverage the free tier where possible, optimize memory allocation to reduce duration, batch events to lower invocations, minimize outbound data transfer by processing data inside the same region, and use cost monitoring to identify high-duration functions. Consider consolidating functions with similar lifecycles and applying cold-start mitigation strategies to reduce unnecessary wake times.