SERPHouse AI SDK

Python SDK for the SERPHouse search API — with first-class integrations for LangChain and LlamaIndex.

Drop live Google, Bing, and Yahoo search data into your AI agents. No boilerplate, no custom API clients — just import, configure, and let your agent search the web.

---

Table of Contents

• Why SERPHouse AI SDK
• Installation
• Quick Start
  • LangChain
  • LlamaIndex
• API Key
• Tools Reference
• Advanced Usage
  • Custom API key per tool
  • SERPHouseClient
• Development
• Contributing
• License

---

Why SERPHouse AI SDK

Zero boilerplate	Two lines to add web search, news, or video results to any LangChain or LlamaIndex agent
Framework-native	Functions return standard StructuredTool / FunctionTool objects — no wrappers, no adapters
Live SERP data	Every call hits the SERPHouse API in real time. No cached or stale results.
Production-ready	Built on requests and pydantic with typed schemas, configurable timeouts, and clear error handling

---

Installation

pip install serphouse-ai-sdk

With framework extras:

pip install serphouse-ai-sdk[langchain]
pip install serphouse-ai-sdk[llamaindex]
pip install serphouse-ai-sdk[all]      # both frameworks

---

Quick Start

Set the SERPHOUSE_API_KEY environment variable (or pass it per tool).

LangChain

from langchain.agents import create_agent
from langchain_openai import ChatOpenAI
from serphouse.langchain import search, news, short_video

agent = create_agent(
    model=ChatOpenAI(model="gpt-4"),
    tools=[search(), news(), short_video()],
)

response = agent.invoke({
    "messages": [{"role": "user", "content": "Latest AI news from OpenAI"}]
})
print(response["messages"][-1].content)

LlamaIndex

from llama_index.core.agent.workflow import FunctionAgent
from llama_index.llms.openai import OpenAI
from serphouse.llamaindex import search, news, short_video

agent = FunctionAgent(
    llm=OpenAI(model="gpt-4"),
    tools=[search(), news(), short_video()],
)

response = agent.run("Latest AI news from OpenAI")
print(response)

---

API Key

Get your key at serphouse.com and choose one of:

Method	Example
Environment variable (recommended)	export SERPHOUSE_API_KEY=your_key_here
Per-tool options dict	search({"api_key": "your_key_here"})
SERPHouseClient	SERPHouseClient(api_key="your_key_here")

---

Tools Reference

Tool	Framework	Endpoint	Description
search()	LangChain, LlamaIndex	/web-search-lite	General web search
news()	LangChain, LlamaIndex	/google-news	Google News search
short_video()	LangChain, LlamaIndex	/google-short-videos-api	Google Short Videos (Shorts)

All tools accept the following parameters:

Parameter	Type	Default	Description
q	str	—	Search query (required)
domain	str	google.com	Search domain
lang	str	en	Search language code
loc	str	New York,New York,United States	Location (City,State,Country)
device	str	desktop	desktop or mobile
page	int	—	Page number
date_range	str	y	h (hour), d (24h), w (week), m (month), y (year), or YYYY-MM-DD,YYYY-MM-DD
gl	str	US	Country code (ISO 3166-1 alpha-2, search only)
video_quality	str	high	Video quality (short_video only)

---

Advanced Usage

Custom API key per tool

from serphouse.langchain import search

tool = search({"api_key": "your-key"})

SERPHouseClient

Use the raw client directly for custom integrations:

from serphouse import SERPHouseClient

client = SERPHouseClient(api_key="your-key")
result = client.post("/web-search-lite", {"q": "python sdk"})
print(result)

---

Development

# Clone and enter the repo
git clone https://github.com/serphouse/ai-sdk.git
cd ai-sdk

# Create a virtual environment
python3 -m venv venv
source venv/bin/activate

# Install in editable mode with all dev dependencies
pip install -e ".[test]"

# Run tests
pytest tests -v

# Run only unit tests (skip OpenAI integration tests)
pytest tests -v -k "not Integration"

# Run integration tests (requires OPENAI_API_KEY)
OPENAI_API_KEY=sk-... pytest tests -v -k "Integration"

---

Contributing

Contributions are welcome. Please keep changes focused and match existing code style.

git checkout -b feature/your-feature
pip install -e ".[test]"
# make changes
pytest tests -v
git commit -m "Add your feature"
git push origin feature/your-feature

Then open a Pull Request. Update this README if you change setup or configuration.

---

License

MIT License — Copyright SERPHouse.