How to use Perplexity AI API as the LLM Model for CrewAI in FastAPI framework.

main.py

import os
from fastapi import FastAPI
from fastapi.responses import JSONResponse
from fastapi.requests import Request
from fastapi.encoders import jsonable_encoder
from typing import List
from pydantic import BaseModel
from crewai import Agent, Task, Crew, Process
from langchain_community.chat_models import ChatPerplexity

# install fastapi and crewai
# set/export the env params for:
# 1. OPENAI_API_KEY
# 2. PPLX_API_KEY
# Run this gist by running `fastapi dev main.py --port {Port Number}`

app = FastAPI()

class Item(BaseModel):
    product: str
    topic: str

class AgentResponse(BaseModel):
    content: str

pplx_client = ChatPerplexity(temperature=0, model="llama-3-sonar-large-32k-chat")

# Define the CrewAI agent
researcher = Agent(
    role='Senior Researcher',
    goal='Get as many information as possible about {product}.',
    verbose=True,
    backstory=(
        "Driven by curiosity, you're at the forefront of"
        "innovation, eager to explore and share knowledge that could change"
        "the world."
    ),
    llm=pplx_client,
)

# Define the CrewAI task
research_task = Task(
    description=(
        "Identify the next big trend in {topic}."
        "Focus on identifying pros and cons and the {product}."
        "Your final report should clearly articulate the key points,"
        "its market opportunities, and potential risks."
    ),
    expected_output='A comprehensive 3 paragraphs long report on the {product}.',
    # tools=[search_tool],
    agent=researcher,
)

# Define the CrewAI crew
crew = Crew(
    agents=[researcher],
    tasks=[research_task],
    process=Process.sequential,
    memory=True,
    cache=True,
    max_rpm=100,
    share_crew=True
)

# Define the route for the Perplexity AI API
@app.post("/research/product")
async def productResearch(item: Item):
    # print(messages)
    response = crew.kickoff(inputs={'product': item.product, 'topic': item.topic})
    return JSONResponse(content={"content": response}, media_type="application/json")