Manage and optimize usage for Vercel KV

The KV section shows the following charts:

The Requests chart shows the total number of Redis commands made to all the KV stores on your Team. Each plan includes a set number of requests. Additional requests made above the included limit in your plan get charged at a set price per unit.

• You can replicate Vercel KV databases in multiple regions to reduce latency and improve availability. However, this means that the number of requests will increase as each write command will be issued once to your primary database and once to each read replicas you have configured
• To disable or stop incurring charges, you should reduce some of your read replicas or stop querying the database

The Data transfer chart shows the amount of data transferred between your KV stores and your compute endpoints. Each plan includes a set amount of data transfer. After that you're charged a set price per unit of additional data transfer made above the included limit in your plan.

• To disable or stop incurring charges, you should reduce some of your read replicas or stop querying the database

The Storage chart shows the maximum amount of data stored on average in all your KV stores per month. Each plan includes a set amount of storage. After that you're charged a set price per unit of additional storage used above the included limit in your plan.

• To reduce cost, you can delete unused data from your KV stores or delete unused stores entirely if they are no longer queried
• Note that read replicas count towards your storage usage, so reducing the number of read replicas can help reduce your storage usage

The Databases chart shows the number of KV databases (including read replicas) that are currently active on your team. Each plan includes 1 KV database. After that, you're charged a set price for each additional database up to the plan maximum. On the Pro plan, you can have a maximum of five databases per month and that each database can have a maximum of five read replicas.

• You get charged based on each database and read replica. You may want to consider deleting read replicas if they are no longer needed