Quickstart

Get started with Airweave in under 10 minutes

Choose your deployment

Before you begin, decide how you want to run Airweave:

Cloud (Recommended)

Managed service with free tier. Get your API key from the dashboard.

Self-hosted

Run locally with Docker. API available at http://localhost:8001

This quickstart uses the cloud deployment. For self-hosted, change base_url to http://localhost:8001 in the examples below.

Install the SDK

Python

pip install airweave-sdk

Step 1: Initialize the client

Get your API key from the Airweave dashboard and initialize the client:

Python

from airweave import AirweaveSDK

# Initialize the client
client = AirweaveSDK(
    api_key="YOUR_API_KEY",
    base_url="https://api.airweave.ai"  # Use "http://localhost:8001" for self-hosted
)

Store your API key in an environment variable: export AIRWEAVE_API_KEY="your-key-here"

Step 2: Create a collection

A collection is a searchable container that groups multiple data sources together. Think of it as a unified search index.

Python

# Create a collection
collection = client.collections.create(
    name="My First Collection"
)

print(f"Created collection: {collection.readable_id}")
print(f"Collection ID: {collection.id}")

# Output:
# Created collection: my-first-collection-x7k9m
# Collection ID: 550e8400-e29b-41d4-a716-446655440000

The readable_id is auto-generated from your collection name with a unique suffix. You’ll use this ID to reference the collection in search queries.

Step 3: Add a source connection

Now connect a data source to your collection. Source connections handle authentication and automatically sync data from your apps and databases.

Stripe (API Key)

Python

# Connect Stripe with API key
source_connection = client.source_connections.create(
    name="My Stripe Connection",
    short_name="stripe",
    readable_collection_id=collection.readable_id,
    authentication={
        "credentials": {
            "api_key": "sk_test_YOUR_STRIPE_API_KEY"
        }
    }
)

print(f"Created: {source_connection.name}")
print(f"Status: {source_connection.status}")

GitHub (API Key)

Python

# Connect GitHub with personal access token
source_connection = client.source_connections.create(
    name="GitHub Docs Repo",
    short_name="github",
    readable_collection_id=collection.readable_id,
    config={
        "repo_name": "airweave-ai/airweave",
        "branch": "main"
    },
    authentication={
        "credentials": {
            "personal_access_token": "ghp_YOUR_TOKEN"
        }
    }
)

print(f"Created: {source_connection.name}")
print(f"Status: {source_connection.status}")

Slack (OAuth)

Python

# Connect Slack with OAuth (browser flow)
source_connection = client.source_connections.create(
    name="Team Slack Workspace",
    short_name="slack",
    readable_collection_id=collection.readable_id,
    redirect_url="https://app.example.com/connections"
)

# For OAuth sources, you'll get an auth_url to complete authentication
print(f"Visit: {source_connection.auth.auth_url}")

Airweave will automatically start syncing data from your source. The initial sync may take a few minutes depending on data volume. You can monitor progress in the dashboard or via the API.

Step 4: Search your collection

Once data is synced, you can search across all connected sources with natural language queries:

Python

# Basic search
results = client.collections.search(
    readable_id=collection.readable_id,
    query="Find recent failed payments"
)

print(f"Found {len(results.results)} results\n")

# Display results
for i, result in enumerate(results.results[:3], 1):
    payload = result['payload']
    print(f"Result {i}:")
    print(f"  Source: {payload['source_name']}")
    print(f"  Content: {payload['md_content'][:100]}...")
    print(f"  Score: {result['score']:.3f}\n")

If you get an error about no results, the initial sync might still be running. Wait a minute and try again, or check the sync status in the dashboard.

Step 5: Try advanced search

Airweave supports powerful search features including filters, AI reranking, and query expansion:

Python

from airweave import SearchRequest, Filter, FieldCondition, MatchAny

# Advanced search with filters and AI features
results = client.collections.search_advanced(
    readable_id=collection.readable_id,
    search_request=SearchRequest(
        query="customer feedback about pricing",
        filter=Filter(
            must=[
                FieldCondition(
                    key="source_name",
                    match=MatchAny(any=["Stripe", "Zendesk", "Slack"])
                )
            ]
        ),
        recency_bias=0.5,        # Prefer newer content
        score_threshold=0.7,     # High-quality results only
        enable_reranking=True,   # AI reranking for better relevance
        limit=10
    )
)

print(f"Found {len(results.results)} high-quality results")

What’s next?

Explore core concepts

Learn about sources, connectors, collections, entities, and syncs

Add more connectors

Connect GitHub, Notion, Slack, databases, and 50+ other sources

Master search features

Filters, reranking, query expansion, and answer generation

Build AI agents

Integrate with Claude, OpenAI, and other AI frameworks

Need help?

Join our Discord

Get help from the community and team

Browse examples

Example notebooks and tutorials