Get started with Exa

1

Set up your API key

Get your API key from the Exa Dashboard and set it as an environment variable.For the Python or JavaScript SDK, create a file called .env in the root of your project and add the following line:

.env

EXA_API_KEY=your api key without quotes

For cURL, set it as an environment variable in your terminal instead:

macOS/Linux
Windows

export EXA_API_KEY="your-api-key"

setx EXA_API_KEY "your-api-key"

2

Install the SDK

Install the SDK for your language. If you want to store your API key in a .env file, also install the dotenv library. cURL needs no installation, skip to the next step.

pip install exa-py
pip install openai
pip install python-dotenv

3

Make your first request

Create a file (exa.py or exa.ts) with the code below, or run the cURL command directly. Pick a use case:

Search and crawl
Answer
Chat Completions

Get a list of results and their full text content.

from exa_py import Exa
from dotenv import load_dotenv

import os

# Use .env to store your API key or paste it directly into the code
load_dotenv()
exa = Exa(api_key=os.getenv('EXA_API_KEY'))

result = exa.search(
  "An article about the state of AGI",
  type="auto",
  contents={"highlights": True}
)

print(result)

Get an answer to a question, grounded by citations from exa.

from exa_py import Exa
from dotenv import load_dotenv

import os

# Use .env to store your API key or paste it directly into the code
load_dotenv()
exa = Exa(api_key=os.getenv('EXA_API_KEY'))

result = exa.stream_answer(
  "What are the latest findings on gut microbiome's influence on mental health?",
  text=True,
)

for chunk in result:
  print(chunk, end='', flush=True)

Get a chat completion from exa.

from openai import OpenAI
from dotenv import load_dotenv

import os

# Use .env to store your API key or paste it directly into the code
load_dotenv()

client = OpenAI(
  base_url="https://api.exa.ai",
  api_key=os.getenv('EXA_API_KEY'),
)

completion = client.chat.completions.create(
  model="exa",
  messages = [
  {"role": "system", "content": "You are a helpful assistant."},
  {"role": "user", "content": "What are the latest developments in quantum computing?"}
],

  extra_body={
    "text": True
  }
)
print(completion.choices[0].message.content)