Amazon Aurora Serverless v2 dynamically adjusts compute capacity to match demand, unlike fixed-size instances. It offers a configurable range for scaling between a minimum and maximum capacity without downtime. Serverless v2 is a type of Aurora instance, inheriting features like distributed storage and Multi-AZ. It utilizes Aurora Capacity Units (ACUs), where one ACU is about 2 GiB of memory with proportional CPU and network. Scaling occurs in-place, preserving active connections and transactions. Users define a minimum and maximum capacity, controlling cost and influencing database parameters. Auto-pause is an option with a minimum capacity of zero, pausing compute to save costs during inactivity, though storage is still charged. Storage for Serverless v2 is the same as conventional Aurora, distributed across Availability Zones. In clusters with readers, promotion tiers dictate how readers scale relative to the writer. Serverless v2 supports a wide range of Aurora ecosystem features including Multi-AZ, read replicas, Global Databases, and RDS Proxy. Its cost model involves ACU-hours for compute and GB-months for storage, with I/O charges depending on the mode. Serverless v2 is ideal for variable workloads, new applications, and dev/test environments, but provisioned instances are often cheaper for consistent high load. Key considerations include setting capacity limits, monitoring ACUs, and optimizing queries rather than solely relying on scaling.