An automated benchmark and public leaderboard for Model-Context Protocol (MCP) clients. This project provides a standardized testing ground to evaluate client capabilities against a consistent set of tasks.
The goal is to foster a robust ecosystem of capable clients by providing developers with a transparent, objective, and easy-to-use evaluation tool.
View the live results, see how different clients perform, and check your own run history on the public dashboard:
➡️ View the Leaderboard at remote-mcp-servers.com/clients
- Automated Testing: A state machine-driven server that guides clients through a series of capability checks.
- Rubric-Based Scoring: A detailed scoring system that awards points based on the successful completion of specific tasks.
- Public Leaderboard: A live, ranked dashboard showcasing the performance of all participating clients.
- Detailed Run Analysis: Drill down into any specific run to see a full scorecard and identify areas for improvement.
- Easy Participation: Simply add our server URL to your favorite MCP client.
Getting your client on the leaderboard is simple.
- Open your preferred MCP client (e.g., Portal One, MCP Inspector).
- Find the option to add a new remote server.
- When prompted for the URL, use the following endpoint:
https://mcp-client-benchmark.remote-mcp-servers.com/mcp
Once connected, the benchmark server will appear in your client. Interacting with it will start the evaluation process. After the run is complete, refresh the leaderboard page to see your results.
Our scoring system is designed to be transparent. Points are awarded based on the successful completion of specific tasks that map to core MCP capabilities.
| Check ID | Description | Max Points |
|---|---|---|
elicitation_support |
Client correctly handles a server-initiated elicitation/request. |
25 |
sampling_support |
Client uses sampling/create_message to generate dynamic text. |
20 |
resource_reading |
Client can read content from a dynamically enabled resource. |
25 |
code_verification |
Client submits data from a resource to a tool correctly. | 25 |
| Total | 95 |
Contributions are welcome! Whether it's improving the benchmark logic, enhancing the dashboard, or fixing a bug, please feel free to open an issue or submit a pull request.
- Fork the repository.
- Create a new feature branch. (
git checkout -b feature/your-amazing-feature) - Commit your changes. (
git commit -m 'Add some amazing feature') - Push to the branch. (
git push origin feature/your-amazing-feature) - Open a Pull Request.
This project is licensed under the MIT License. See the LICENSE file for details.