谷歌推出 File Search Tool 文件搜索系统 ，集成至 Gemini API

from google import genai
from google.genai import types
<p>client = genai.Client()
store = client.file_search_stores.create()</p><p>upload_op = client.file_search_stores.upload_to_file_search_store(
file_search_store_name=store.name,
file='path/to/your/document.pdf'
)</p><p>while not upload_op.done:
time.sleep(5)
upload_ops = client.operations.get(upload_op)</p><h1>在生成请求中将文件搜索存储作为工具使用</h1><p>response = client.models.generate_content(
model='gemini-2.5-flash',
contents='What does the research say about ...',
config=types.GenerateContentConfig(
tools=[types.Tool(
file_search=types.FileSearch(
file_search_store_names=[store.name]
)
)]
)
)</p><p>print(response.text)</p>
                    <div class="aritcle_card">
                        <a class="aritcle_card_img" href="/ai/1385">
                            <img src="https://img.php.cn/upload/ai_manual/000/000/000/175680265595886.png" alt="纳米搜索">
                        </a>
                        <div class="aritcle_card_info">
                            <a href="/ai/1385">纳米搜索</a>
                            <p>纳米搜索：360推出的新一代AI搜索引擎</p>
                            <div class="">
                                <img src="/static/images/card_xiazai.png" alt="纳米搜索">
                                <span>30</span>
                            </div>
                        </div>
                        <a href="/ai/1385" class="aritcle_card_btn">
                            <span>查看详情</span>
                            <img src="/static/images/cardxiayige-3.png" alt="纳米搜索">
                        </a>
                    </div>
                <h1>使用来源链接支持回答内容</h1><p>grounding = response.candidates[0].grounding_metadata
sources = {c.retrieved_context.title for c in grounding.grounding_chunks}
print('Sources:', *sources)