采样能解锁什么
让服务器将推理工作委托回模型。
采样能解锁什么 是 CoddyKit 上的免费 MCP Academy 课时。 这是第 1 节课，共 4 节。 你可以在下方免费阅读本课时的完整内容 — 然后在浏览器中使用内置代码编辑器和全天候 AI 导师进行实践。 这是 MCP Academy 学习路径的一部分，你的进度在网页和 CoddyKit 应用中同步。 MCP Academy 课程共包含 4 节课。
本课时的部分内容尚未翻译，以英文显示。
Flipping the Direction
Usually the model calls your tools. Sampling flips that: your server asks the model to generate text for it. 🔄
Borrow the Model You Already Have
With sampling, your server can use the host’s LLM without ever holding an API key. The client owns the model access.
No API Keys in Your Server
Because the client runs the completion, your server needs no provider credentials. That keeps secrets out of your code entirely.
Agentic Behavior, Nested
Sampling lets an LLM call happen nested inside a tool. Your tool can think with the model mid-task, not just return data.
A Tiny Example: Summarize
Imagine a tool that fetches a report, then asks the model to summarize it. The summarizing happens through sampling.
You Stay in Control of Cost
The client picks the model and pays for it, so the host decides cost, limits, and which model runs your request.
Server Delegates Reasoning
Think of sampling as the server saying: I have data, you have a brain. It delegates the reasoning step back to the model.
What You Can Ask For
A sampling request can ask the model for text, and clients may also support image or audio content in messages.
It Is Opt-In on the Client
Sampling only works if the client declares the sampling capability at startup. Not every host enables it.
"capabilities": {
"sampling": {}
}Why It Matters
Sampling unlocks smart, self-contained servers: data tools that can reason, classify, and write — powered by the user’s own model.
The Catch: Trust
Because servers can trigger model calls, there must be a human able to approve them. You’ll see that safety flow soon.
Quick Check
Quick gut check on what sampling really does.
Recap
Sampling flips control: your server asks the host’s model to think for it — no keys, opt-in, and always with a human able to approve. ✅
常见问题解答
「采样能解锁什么」课时是免费的吗？
是的 — 「采样能解锁什么」的完整文本可在网页上免费阅读。要进行交互式练习（内置代码编辑器和全天候 AI 导师）并解锁 MCP Academy 课程的其余内容，请升级到 CoddyKit PRO。 MCP Academy 课程共包含 4 节课。
「采样能解锁什么」这节课中我会学到什么？
让服务器将推理工作委托回模型。 你通过在浏览器中直接运行的动手代码来练习 MCP Academy，全天候 AI 导师会在你学习这节课的过程中回答你的问题。
学习 MCP Academy 需要有经验吗？
无需任何先前经验。CoddyKit 上的 MCP Academy 课程适合初学者到高级学习者，你可以从这里开始或从头开始，按照自己的节奏学习。 这是第 1 节课，共 4 节。
「采样能解锁什么」课时需要多长时间？
大多数 CoddyKit 课程大约需要 5–10 分钟。每节课都很精短且互动，所以你能稳步进步，并在网页和应用中从离开的地方继续。
我能在这节 MCP Academy 课中编写并运行代码吗？
能。每节 MCP Academy 课都包含内置代码编辑器，你可以在浏览器中直接编写并运行真实代码，并获得即时 AI 反馈 — 无需本地设置。