AWS Lambda is a serverless compute service provided by Amazon Web Services (AWS). It allows you to run your code without provisioning or managing servers. The pricing for AWS Lambda is based on a pay-per-use model, primarily determined by two factors:

1. Number of Requests: AWS Lambda charges based on the number of requests made to your functions. This includes both invocation requests and event source mapping requests. The pricing is tiered, meaning the cost per request decreases as the number of requests increases.

2. Duration of Execution: AWS Lambda charges based on the duration of the execution of your function code, measured in increments of 100 milliseconds. You are billed for the total execution time, including any initialization and shutdown phases.

There are a few additional considerations and pricing details to note:

1. Free Tier: AWS provides a free tier for AWS Lambda that includes a certain number of free requests and free compute time per month. This allows you to experiment and get started with Lambda at no cost.

2. Compute Time: AWS Lambda bills compute time in increments of 100 milliseconds, rounding up to the nearest increment. So, even if your function runs for a shorter duration, you will still be billed for the minimum increment of 100 milliseconds.

3. Memory Allocation: You can choose the amount of memory allocated to your function, ranging from 128 MB to 10 GB. The pricing is based on the allocated memory, and higher memory configurations have a higher cost per 100 milliseconds of execution time.

4. Additional Services and Resources: If your AWS Lambda function interacts with other AWS services or resources, such as Amazon S3, Amazon DynamoDB, or Amazon RDS, additional costs may apply based on the usage and pricing of those services.