Side-by-side MCP server comparison
Short answer: it is close. Agnt scores 93/100 and io.github.zenml-io/mcp-zenml scores 93/100 - a 0-point gap, which is inside the noise of a weekly rescore. Agnt exposes 7 tools against io.github.zenml-io/mcp-zenml's 17, so the choice comes down to whether you need that extra surface area or would rather keep your context window small.
Last updated
| Metric | Agnt | io.github.zenml-io/mcp-zenml |
|---|---|---|
| MCPpedia Score | 93 | 93 |
| Security Score | 30 | 30 |
| Maintenance Score | 18 | 18 |
| GitHub Stars | 22 | 49 |
| Weekly Downloads | 2263.3k | 98.7k |
| Tools | 7 | 17 |
| Est. Token Cost | 1.1k | 2.5k |
| Open CVEs | 0 | 0 |
| Health | Active | Active |
| Auth Required | No | No |
| Transport | stdio, sse, http | stdio, sse, http |
| License | Apache-2.0 | MIT |
| API Pricing | unknown | freemium |
Agnt and io.github.zenml-io/mcp-zenml are closely matched at the top with scores of 93 and 93. Check each server's detail page to decide which fits your use case.
Compare up to 4 servers: mcppedia.org/compare/server-a-vs-server-b-vs-server-c