Build AI Agents with RapidAPI for Real-Time Data

Creating brokers has by no means been simpler, however have you ever ever questioned how one can make them much more highly effective than they’re now?I not too long ago considered one doable methodology. What when you had real-time details about particular classes like finance or films? That is actually nice, is not it? Whereas contemplating this selection, I found RapidAPI, a hub of APIs that permits you to give your AI brokers entry to real-time info with a easy API name. Subsequent, I made a decision to create some brokers that may reap the benefits of these APIs to create higher role-playing brokers.

On this article, I’ll share the complete course of so you may simply comply with it and reproduce the outcomes to your personal use. Let’s begin with some fundamental info –

What’s RapidAPI?

RapidAPI is definitely an outdated identify and not too long ago turned Nokia API Hub after an acquisition. There’s a catalog of APIs that you should utilize or expose. It covers numerous classes reminiscent of cybersecurity, films, communication, and so forth. Be taught extra about RapidAPI.

How do I exploit RapidAPI’s API?

1. First check in/join RapidAPI right here.

2. Go to the Developer Authentication web page and click on Add Authentication within the prime proper nook to create an authentication of sort RapidAPI.

3. Return to the house web page, discover APIs, and click on on any API. For instance, right here I clicked on the Cryptocurrency Information API.

4. A web page like this can be displayed. The API key can be already current within the check code. Be sure that the goal is ready to “Python”.

5. Subsequent, on the highest proper[テストに登録]Click on and choose the free tier for now. Then click on “Begin Free Plan” after which click on Subscribe.

6. Now you should utilize the check endpoint button within the prime proper nook to run your check code and get the response.

Be aware: Most APIs have a beneficiant free tier that permits you to use them as much as the month-to-month limits listed.

Combine the agent with RapidAPI

On this part, we’ll create an agent utilizing the “create_agent” perform in langchain.brokers. The agent makes use of OpenAI, particularly “gpt-5-mini”. Be happy to experiment with totally different fashions, mannequin suppliers, and agent frameworks.

Conditions

To keep away from repetition, use the identical set of imports and initialize the API in order that it may be utilized by a number of brokers. If you wish to check it with me, you’ll want to subscribe to the API within the hyperlink. We additionally use Google Colab for the demo.

Subscribe to those APIs

Configure a Google Colab pocket book

Do not forget so as to add OpenAI API and RapidAPI to the secrets and techniques part on the left as “OPENAI_API_KEY” and “RAPIDAPI_KEY” and activate entry to your pocket book.

set up

!pip set up langchain langchain_core langchain_openai -q

Imported items

Import userdata from google.colab Import os Import json from http.shopper Import instruments from langchain_core.instruments Import from langchain_openai ChatOpenAI Import from langchain.brokers create_agent

API key

OS.Atmosphere[‘OPENAI_API_KEY’] = userdata.get(‘OPENAI_API_KEY’) RAPIDAPI_KEY = userdata.get(‘RAPIDAPI_KEY’)

Constructing a information agent

@instrument def search_news(question: str, restrict: int = 10) -> str: “””Seek for real-time information articles primarily based on a question. Returns the newest information articles.””” conn = http.shopper.HTTPSConnection(“real-time-news-data.p.rapidapi.com”) headers = { “x-rapidapi-key”: RAPIDAPI_KEY, “x-rapidapi-host”: “real-time-news-data.p.rapidapi.com”, } conn.request( “GET”, f”/search?question={question}&restrict={restrict}&time_published=anytime&nation=US&lang=en”, headers=headers, ) res = conn.getresponse() knowledge = res.learn() consequence = json.hundreds(knowledge.decode(“utf-8″)) return json.dumps(consequence, indent=2) news_agent = create_agent( ChatOpenAI(temperature=0, mannequin=”gpt-5-mini”), instruments=[search_news],)

Be aware that we’re utilizing the API offered by RapidAPI as a instrument and passing that instrument to the agent. The agent enlists the assistance of the instrument every time it feels it must be invoked.

# Take a look at the agent consequence = news_agent.invoke({ “messages”: [{“role”: “user”, “content”: “Search for latest news about Messi”}]

}) print(consequence[“messages”][-1]. content material)

consequence

fantastic! I’ve created my first agent and it is going properly. You’ll be able to check out new prompts when you like.

Be aware: Our brokers primarily reply to questions with one phrase (e.g. “sports activities”, “forest”, and so forth.). It is because the instrument solely accepts single strings, not sentences. To repair this, configure the agent’s system immediate.

inventory company

Let’s use Yahoo’s API to create a inventory agent that retrieves inventory particulars utilizing the inventory ticker image for a specific inventory.

code

@instrument def get_stock_history(image: str, Interval: str = “1m”, restrict: int = 640) -> str: “””Will get the historic inventory worth knowledge of the image.””” conn = http.shopper.HTTPSConnection(“yahoo-finance15.p.rapidapi.com”) headers = { “x-rapidapi-key”: RAPIDAPI_KEY, “x-rapidapi-host”: “yahoo-finance15.p.rapidapi.com”, } path = ( f”/api/v2/markets/inventory/historical past?” f”image={image}&interval={interval}&restrict={restrict}” ) conn.request(“GET”, path, headers=headers) res = conn.getresponse() knowledge = res.learn() consequence = json.hundreds(knowledge.decode(“utf-8″)) return json.dumps(consequence, indent=2) Stock_agent = create_agent( ChatOpenAI(temperature=0, mannequin=”gpt-5-mini”), instruments=[get_stock_history]) # Instance name consequence = Stock_agent.invoke( {“messages”: [{“role”: “user”, “content”: “Get the last intraday price history for AAPL”}]} ) print(consequence[“messages”][-1]. content material)

consequence

I used to be capable of efficiently acquire the output of AAPL (Apple Inc.). Data is totally real-time.

property agent

Our aim right here is to create an agent with an API that searches for properties on the market/lease. The API utilized by Zoopla particularly searches for properties inside the UK.

code

@instrument def search_properties( location_value: str, location_identifier: str = “metropolis”, web page: int = 1, ) -> str: “”” Seek for residential properties. Arguments: location_value: Title of location (e.g. ‘London’, ‘Manchester’, ‘E1 6AN’). location_identifier: Class of location. – Use ‘metropolis’ for main cities (default). – If the consumer specifies a postal code, use ‘postal_code’ (e.g. ‘W1’) – use ‘space’ for smaller areas. “”” # URL encoding to stop InvalidURL errors. http.shopper.HTTPSConnection(“zoopla.p.rapidapi.com”) headers = { “x-rapidapi-key”: RAPIDAPI_KEY, “x-rapidapi-host”: “zoopla.p.rapidapi.com”, } path = ( f”/properties/v2/record?locationValue={safe_val}” f”&locationIdentifier={safe_id}” f”&class=residential&furnishedState=Any” f”&sortOrder=newest_listings&web page={web page}” ) conn.request(“GET”, path, headers=headers) res = conn.getresponse() knowledge = res.learn() consequence = json.hundreds(knowledge.decode(“utf-8”)) return json.dumps(consequence, indent=2) property_agent = create_agent( ChatOpenAI(temperature=0, mannequin=”gpt-5-mini”), instruments=[search_properties]system_prompt=( “You’re a actual property skilled. When a consumer asks for a location,” “Guess the “location_identifier” your self. Normally “metropolis” or “postal_code.” Do not ask the consumer for a technical identifier.” “Invoke the instrument now.” ), ) consequence = property_agent.invoke( {“messages”: [{“role”: “user”, “content”: “Search for properties in London, England”}]} ) print(consequence[“messages”][-1]. content material)

consequence

I acquired the precise properties as output, however they’re blurred for apparent causes.

film suggestion agent

The agent now has entry to each the very best and lowest rated APIs on IMDB as a instrument, and configures system prompts to pick out which instrument to make use of primarily based on the immediate.

@instrument def get_top_rated_movies() -> str: “”” Get the record of prime rated English films on IMDb. Use this when customers need suggestions or “good” films. } conn.request(“GET”, “/api/imdb/top-rated-english-movies”, headers=headers) res = conn.getresponse() # Decode and return the uncooked JSON for agent processing return res.learn().decode(“utf-8”) @instrument def get_lowest_rated_movies() -> str: “”” Get the record of lowest rated films on IMDb. Use this when a consumer requests “unhealthy” films or films to keep away from. } conn.request(“GET”, “/api/imdb/lowest-rated-movies”, headers=headers) res = conn.getresponse() return res.learn().decode(“utf-8″) movie_agent = create_agent( ChatOpenAI(temperature=0, mannequin=”gpt-5-mini”), instruments=[get_top_rated_movies, get_lowest_rated_movies]system_prompt=( “You’re a skilled film critic. Your aim is to assist customers discover films primarily based on high quality. If the consumer requests “good,” “beneficial,” “ “ or “basic,” name get_top_rated_movies. Use the parameters to concisely summarize the ends in one sentence: ” ), ) # Instance utilization consequence = movie_agent.invoke( { “messages”: [
{
“role”: “user”,
“content”: “I’m in the mood for a really terrible movie, what’s the worst out there?”,
}
]
} ) print(consequence[“messages”][-1]. content material)

consequence

If you would like one thing actually unhealthy, IMDb’s lowest-rated titles embody Daniel der Sauberer (Daniel the Wizard) and Smolensk. Each have a mean score of round 1.2, making them excellent when you’re searching for one thing “so unhealthy, so fascinating.”

fantastic! We had been capable of create an agent that has entry to a number of instruments and might recommend each the very best and lowest rated films.

conclusion

By integrating real-time APIs with brokers, you may transcend static responses and construct methods that really feel really clever. RapidAPI makes this integration easy and scalable throughout domains. It is also vital to decide on the correct instruments and tune your brokers to work in concord with these instruments. For instance, many APIs may cause errors if their arguments comprise single quotes or areas. RapidAPI additionally offers MCP help all through the API, which you’ll be able to think about in your continued efforts to create higher brokers.

FAQ