> ## Documentation Index
> Fetch the complete documentation index at: https://docs-vip.apigo.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Gemini 思考模式示例

> Gemini thinking budget 接入示例。

## 推荐 endpoint

* [Gemini /v1beta/models/{model}:generateContent](/api-reference/endpoints/gemini/text)

## 最小请求

```json theme={null}
{
  "model": "gemini-2.5-pro",
  "contents": [
    {
      "role": "user",
      "parts": [{ "text": "比较三种缓存架构的取舍，并给出推荐。" }]
    }
  ],
  "generationConfig": {
    "thinkingConfig": {
      "thinkingBudget": 1024
    }
  }
}
```

## cURL 示例

```bash theme={null}
curl "https://generativelanguage.googleapis.com/v1beta/models/gemini-2.5-pro:generateContent" \
  -H "x-goog-api-key: $GEMINI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "contents": [
      {
        "role": "user",
        "parts": [{ "text": "比较三种缓存架构的取舍，并给出推荐。" }]
      }
    ],
    "generationConfig": {
      "thinkingConfig": {
        "thinkingBudget": 1024
      }
    }
  }'
```

## Python 示例

```python theme={null}
import requests

response = requests.post(
    "https://generativelanguage.googleapis.com/v1beta/models/gemini-2.5-pro:generateContent",
    headers={
        "x-goog-api-key": "<GEMINI_API_KEY>",
        "Content-Type": "application/json",
    },
    json={
        "contents": [
            {
                "role": "user",
                "parts": [{"text": "比较三种缓存架构的取舍，并给出推荐。"}],
            }
        ],
        "generationConfig": {
            "thinkingConfig": {
                "thinkingBudget": 1024,
            }
        },
    },
    timeout=60,
)
response.raise_for_status()

print(response.json()["candidates"][0]["content"]["parts"][0]["text"])
```

## Node.js 示例

```js theme={null}
const response = await fetch(
  "https://generativelanguage.googleapis.com/v1beta/models/gemini-2.5-pro:generateContent",
  {
    method: "POST",
    headers: {
      "x-goog-api-key": process.env.GEMINI_API_KEY,
      "Content-Type": "application/json"
    },
    body: JSON.stringify({
      contents: [
        {
          role: "user",
          parts: [{ text: "比较三种缓存架构的取舍，并给出推荐。" }]
        }
      ],
      generationConfig: {
        thinkingConfig: {
          thinkingBudget: 1024
        }
      }
    })
  }
);

const data = await response.json();
console.log(data.candidates[0].content.parts[0].text);
```

## 最佳实践

* 高质量推理优先用 Pro 类模型
* `thinkingBudget` 要按任务复杂度逐步加，不要一开始拉满
* 对低延迟路径保留一个 `thinkingBudget: 0` 的降级配置