I'm considering using Weaviate Cloud Service (WCS) for my project, which requires a scalable vector search engine. Given the nature of my project, I anticipate the need to scale horizontally, potentially resulting in a significant number of clusters. Before proceeding, I need to understand the cost implications of such a setup within WCS.
Could anyone provide insights into how WCS's pricing model works, especially in relation to the following points?
- Number of Clusters: Is the pricing affected by the number of clusters deployed? If so, how does WCS calculate the cost related to clusters?
- Resource Usage: How does WCS charge for the computational resources (e.g., CPU, memory) and the volume of data stored? Is there a direct correlation between the number of clusters and resource usage in terms of pricing?
- Query Volume: Does the number of queries or operations performed impact the overall cost, and is this linked in any way to the cluster count?
- Additional Features: Are there any costs associated with scaling, such as using additional features or support options that become necessary with more clusters?
Any detailed information or experiences with WCS, especially regarding scalability and its cost implications, would be greatly appreciated.
Did you see the pricing page for WCS?
On the page, you can find a calculator that will help you estimate the cost of running Weaviate Serverless.
The pricing is largely tied to the
number of dimensionsfor your data. Thenumber of dimensionscan be calculated by multiplying theobject count(of those with vectors) byvector dimensionality.For example, if you have 100'000 objects with 1536-dimension vectors. Then the number of dimensions is 100000 x 1536 = 153'600'000
Overall, WCS will automatically adjust to your needs, so you shouldn't need to worry about scaling the underlying resources.
I hope this helps.