许多LangChain代理商使用BraveSearchResults作为他们的网络搜索工具。随着 Brave 的免费套餐于 2026 年 2 月取消，这些代理商现在面临着意想不到的成本。本教程将 Brave 工具替换为 langchain-Scavio，每月提供 250 个免费积分并可访问 5 个平台。
前置条件
- Python 3.10+
- 浪链已安装
- Scavio API 密钥
操作指南
步骤 1: 安装 langchain-scavio
将 Brave 搜索包替换为 Scavio 的 LangChain 集成。
Bash
# Remove Brave search (optional, but clean):
pip uninstall langchain-community # if only used for Brave
# Install Scavio LangChain integration:
pip install langchain-scavio步骤 2: 替换工具初始化
在代理设置中将 BraveSearchResults 替换为 ScavioSearchResults。
Python
# BEFORE (Brave):
# from langchain_community.tools import BraveSearch
# search_tool = BraveSearch.from_api_key(api_key=os.environ['BRAVE_API_KEY'])
# AFTER (Scavio):
from langchain_scavio import ScavioSearchResults
import os
search_tool = ScavioSearchResults(
api_key=os.environ['SCAVIO_API_KEY'],
platform='google', # or 'reddit', 'youtube', 'amazon', 'walmart'
)步骤 3: 添加多平台工具（可选）
添加特定于平台的工具以获得更丰富的代理功能。
Python
from langchain_scavio import ScavioSearchResults
# Create tools for different platforms:
google_tool = ScavioSearchResults(api_key=os.environ['SCAVIO_API_KEY'], platform='google')
reddit_tool = ScavioSearchResults(api_key=os.environ['SCAVIO_API_KEY'], platform='reddit')
youtube_tool = ScavioSearchResults(api_key=os.environ['SCAVIO_API_KEY'], platform='youtube')
# Use in agent:
tools = [google_tool, reddit_tool, youtube_tool]
# The agent can now search Google for facts, Reddit for opinions,
# and YouTube for tutorials - all under one API key and credit pool.步骤 4: 更新您的代理
将新工具连接到您现有的 LangChain 代理中。
Python
from langchain.agents import create_tool_calling_agent, AgentExecutor
from langchain_anthropic import ChatAnthropic
from langchain_core.prompts import ChatPromptTemplate
llm = ChatAnthropic(model='claude-sonnet-4-6')
tools = [google_tool, reddit_tool]
prompt = ChatPromptTemplate.from_messages([
('system', 'You are a research assistant. Use search tools to find current information.'),
('human', '{input}'),
('placeholder', '{agent_scratchpad}'),
])
agent = create_tool_calling_agent(llm, tools, prompt)
executor = AgentExecutor(agent=agent, tools=tools)
result = executor.invoke({'input': 'What is the current state of the SerpAPI lawsuit?'})
print(result['output'])Python 示例
Python
from langchain_scavio import ScavioSearchResults
import os
tool = ScavioSearchResults(api_key=os.environ['SCAVIO_API_KEY'], platform='google')
result = tool.invoke('python web framework comparison 2026')
print(result)JavaScript 示例
JavaScript
// langchain-scavio is Python-only. For JS, use HTTP directly:
const tool = {
name: 'web_search',
func: async (query) => {
const r = await fetch('https://api.scavio.dev/api/v1/search', {
method: 'POST', headers: {'x-api-key': process.env.SCAVIO_API_KEY, 'Content-Type': 'application/json'},
body: JSON.stringify({platform: 'google', query})
});
return JSON.stringify((await r.json()).organic?.slice(0,5));
}
};预期输出
JSON
A LangChain agent using langchain-scavio instead of BraveSearch, with multi-platform search capabilities and 250 free credits per month.