Compare MCP servers
Put two or three servers side by side across every field in the catalogue. Rows where the servers differ are highlighted, and identical rows can be hidden.
Choose servers
Pick at least two servers to compare.
Comparison
| Field | Server 1 | Server 2 | Server 3 |
|---|---|---|---|
| Description | |||
| Publisher | |||
| Registry name | |||
| Version | |||
| Category | |||
| Distribution | |||
| Transports | |||
| Package type | |||
| Package identifier | |||
| Runtime | |||
| Hosted endpoint | |||
| Authentication | |||
| Credentials required | |||
| Required variables | |||
| Source repository | |||
| Repository status | |||
| Website | |||
| Publisher namespace | |||
| Registry status | |||
| First published | |||
| Registry updated | |||
| Schema generation | |||
| Catalogue score ? |
Nothing selected yet. Search above, or use the compare checkboxes on any server list.
- Yellow row label
- The servers answer this field differently. One offers something the other does not, or they made opposite choices. These are the rows that decide which server suits you.
- Grey row label
- Every server shown gives the same answer, so this field cannot separate them. Worth skipping unless you want to confirm a shared requirement, such as all of them needing an API key.
- A hyphen in a cell
- There is no value for that server. It means either the publisher did not disclose it, or the check that would fill it has not run yet. Those are different things, so open the server page, which states which of the two applies.
- Catalogue score ?
- Points out of 100 for how well documented and verifiable a server is. It rewards a published repository, a tool manifest, a verified publisher domain and recent activity. It says nothing about whether the server is any good at its job. Hover the question mark for the full formula.