MCPJam is the first testing & evaluations platform for MCP servers. Over 106,000 developers and +300 enterprises use MCPJam to see how their servers behave in apps like ChatGPT and Claude, improve experiences for users in AI-native chat interfaces, and build continuous reliability into every deployment.
Headquartered in San Francisco, MCPJam aims to help developers build better MCP servers.
This is the 2nd launch from MCPJam . View more
MCPJam
Launched this week
Run User Testing, Swarms, Evals, and CI/CD gates on your MCP server to see if users actually succeed in ChatGPT, Claude, and Copilot. Test local servers via desktop app, CLI, or SDK.
Free
Launch Team / Built With
Postman
MCPJam
Hey Product Hunt 👋 Prathmesh, CEO of MCPJam here.
Users now start in ChatGPT, Claude, Cursor, and other AI clients. They reach your product through your MCP server.
That means your users often aren’t in your product anymore. You can’t see what they prompted for, how the agent interpreted it, or whether your server helped them get the result they wanted.
I saw this firsthand leading MCP technical strategy at Asana, including our ChatGPT and Claude launches. We were building high-stakes enterprise integrations, but we had no reliable way to test them the way we test normal software- or to know whether they worked once they reached real users.
I started using MCPJam for those problems after re-connecting with my former coworker who created the project. brought it to more of our developers, and worked it into our CI/CD pipeline. I joined the team because I kept hearing the same issue from other companies building for agents.
So, what does “good” look like for MCP? For us, it means users reliably get the outcome they came for, across the AI clients they use.
That’s what we’ve been building toward. MCPJam now helps you test the full workflow, from the first prompt to the expected result:
* Swarms: Simulate users with different goals and prompts to find where workflows break across AI clients.
* User Testing: Watch how real users interact with your MCP product, where they get stuck, and how they feel about the results.
* Evals: Turn those workflows into repeatable tests that check whether users get the expected outcome.
* CI/CD: Run those evals across AI clients before each release to catch regressions.
MCPJam has grown from a debugging tool into a continuous testing and evaluation workflow for MCP servers.
If you’re building an MCP server or agent-facing product, give MCPJam a try. What is the hardest thing for you to test? We love hearing about your MCP server builds!
The amount of times ChatGPT/Codex/Claude have changed how they deal with plugins/MCP's etc. is insane (every few days it keeps changing), but it's been impressive how MCP Jam is able to stay on top of the latest changes and let us know how our MCP will interact in these different apps/scenarios.
Because we've definitely tried to stay on top of it without it, and it's been a mess.
Do you have ways of also helping test/evaluate different access levels to tools? In case we want to create actors that access our MCP in cloud agents that have a restricted set of scopes/permissions?
Congrats on the launch team!
MCPJam
@brian_kim21 great to hear and thanks! Absolutely- mixed auth, step-up authorization, and scoped oauth access generally you can test now on our platform, alongside our OAuth and XAA / EMA debugger!
Clueso
This is super cool! We've actually struggled a bit with answering questions like 'How many tokens will this workflow consume?' when we're discussing our MCP with procurement teams. Is this something you can help with? Would be awesome to get to see a given prompt/workflow and the net cost with the MCP across clients, and benchmark those.
MCPJam
@prajwal_prakash Yeah definitely man! You can run a prompt side-by-side across a bunch of major AI clients and see the full trace and the exact input and output tokens at each step in our playground, then do the same in evals test cases. Check it out!
Clueso
@prathmesh_patel Awesome, will that give a dollar value too? That'd be superb
MCPJam
@prajwal_prakash Absolutely- able to with our tokenizer framework- yes!
Sounds cool... But how is this different from Braintrust?
MCPJam
@mioo101 great question! Braintrust and platforms like it help you evaluate your own agents. They evaluate your agent's behavior and sit inside your own application. MCPJam helps you evaluate how your server behaves inside other clients like ChatGPT, Claude, Cursor, etc. that you don't control.
Is there a way to automate evaluations instead of manually testing each tool?
What kind of tests do you recommend running before pushing an MCP server into production?
Looks great! Particularly excited for evals that can run for different clients. I’m often dealing with customers using gemini, chatgpt, claude etc asking why their prompts aren’t doing what they expect when communicating with our MCP server, and evals will help bridge that gap