The key difference
Client-side encryption happens before the data leaves your device. Server-side encryption happens after the server has already received the plaintext. Those two models can both use strong algorithms, but they do not create the same privacy guarantees.
If the server sees the plaintext, then the platform, its logs, and its operators have a chance to encounter the secret. That may be acceptable for some workloads, but it is not ideal for passwords, API keys, or private notes.
Choosing the right model
Client-side encryption is the better fit when the goal is zero-knowledge sharing. It is especially useful when you want the platform to act as a courier rather than a vault.
Server-side encryption still has its place for many business systems, but it should not be confused with end-to-end privacy. The moment plaintext reaches the server, the trust boundary has already expanded.