Python SDK

​

Available Models

• Real-Time Search (AI Models): Perform live web and financial queries using natural language.
• AI Recommendations (Data Models): Retrieve curated, vertical-specific content for domains like pets, lifestyle, and news.

​

🔎 Real-Time Search (AI Models)

​

🤖 AI Recommendations (Data Models)

​

Installation

pip install dappier

​

Initialization

import os
from dappier import Dappier

# Set your API key as an environment variable
os.environ["DAPPIER_API_KEY"] = "<YOUR_API_KEY>"

# Initialize the Dappier SDK
app = Dappier()

​

Real-Time Search

response = app.search_real_time_data(
    query="What is the stock price of Apple?",
    ai_model_id="am_01j06ytn18ejftedz6dyhz2b15"
)

​

Parameters (Real Time Search)

​

query (string):

• A natural language query representing the information being searched for in real time.

​

ai_model_id (string):

• The ID of the AI model to be used for real-time data search.
• This must be a valid model ID from the Dappier Marketplace.

​

AI Recommendations

response = app.get_ai_recommendations(
    query="latest trending news",
    data_model_id="dm_01jagy9nqaeer9hxx8z1sk1jx6",
    similarity_top_k=5,
    ref="techcrunch.com",
    num_articles_ref=2,
    search_algorithm="most_recent"
)

​

Parameters (AI Recommendations)

​

query (string):

• A natural language query or URL.

​

data_model_id (string):

• The ID of the data model to be used for recommendations.
• This must be a valid model ID from the Dappier Marketplace.

​

similarity_top_k (integer):

• The number of articles to return (default is 9).

​

ref (string):

• The domain of the site from which the recommendations should come.
• Example: techcrunch.com.

​

num_articles_ref (integer):

• Specifies how many articles should be guaranteed to match the domain specified in ref.
• Use this to ensure a set number of articles from the desired domain appear in the results.

​

search_algorithm (string):

• Options: "most_recent" or "semantic".
• "semantic" (default): Contextual matching of the query to retrieve articles.
• "most_recent": Retrieves articles sorted by the most recent publication date.

​

Async Functionality

​

Async Real Time Search:

import asyncio
from dappier import DappierAsync

async def fetch_real_time_data(query):
    dp_client = DappierAsync(api_key=os.environ["DAPPIER_API_KEY"])
    async with dp_client as client:
        response = await client.search_real_time_data_async(
            query=query,
            ai_model_id="am_01j06ytn18ejftedz6dyhz2b15"
        )

        if response is None:
            raise Exception("An error occurred while retrieving the response.")
        return response.message

query = "What is the stock price of Apple?"
result = asyncio.run(fetch_real_time_data(query))
print(result)

​

Parameters (Real Time Search)

​

query (string):

• A natural language query representing the information being searched for in real time.

​

ai_model_id (string):

• The ID of the AI model to be used for real-time data search.
• This must be a valid model ID from the Dappier Marketplace.

​

Async AI Recommendations:

import asyncio
from dappier import DappierAsync

async def fetch_ai_recommendations(query):
    dp_client = DappierAsync(api_key=os.environ["DAPPIER_API_KEY"])
    async with dp_client as client:
        response = await client.get_ai_recommendations_async(
            query=query,
            data_model_id="dm_01jagy9nqaeer9hxx8z1sk1jx6",
            similarity_top_k=5,
            ref="techcrunch.com",
            num_articles_ref=2,
            search_algorithm="most_recent"
        )

        if response is None or response.status != "success":
            raise Exception("An error occurred while retrieving the response.")

        return response.results

query = "latest trending news"
result = asyncio.run(fetch_ai_recommendations(query))
print(result)

​

Parameters (AI Recommendations)

​

query (string):

• A natural language query or URL.

​

data_model_id (string):

• The ID of the data model to be used for recommendations.
• This must be a valid model ID from the Dappier Marketplace.

​

similarity_top_k (integer):

• The number of articles to return (default is 9).

​

ref (string):

• The domain of the site from which the recommendations should come.
• Example: techcrunch.com.

​

num_articles_ref (integer):

• Specifies how many articles should be guaranteed to match the domain specified in ref.
• Use this to ensure a set number of articles from the desired domain appear in the results.

​

search_algorithm (string):

• Options: "most_recent" or "semantic".
• "semantic" (default): Contextual matching of the query to retrieve articles.
• "most_recent": Retrieves articles sorted by the most recent publication date.