An End-to-End RAG Toolkit

SciPhi is an open source framework for testing, deploying,
and monitoring LLM chat integrations with your data and the web

Docs Discovery Call

API Quick Start

Setup Instructions

Python API

# Note, more functionality is coming soon!

# Requires SCIPHI_API_KEY in the environment
from agent_search import SciPhi

client = SciPhi()

# Search the web, then summarize result and generate related queries
agent_summary = client.agent_search_summary(
      query='latest news',
      search_provider='bing',
      llm_model='SciPhi/Sensei-7B-V1'
)

print(agent_summary)
# {
#    "response": "The latest news encompasses ... and its consequences [2].",
#    "related_queries": ['Details on the...', ...],
#    "search_results": [...]
# }

AgentSearch

Building Open Source Web-Scale Search

Join our initiative to enable LLM agents with seamless access to humanity's key knowledge.

Web-Scale Search

Leverage our web-scale search to ground LLM outputs and generate synthetic data, empowering experts to train superior models efficiently.