Framework Integrations

Airweave provides native integrations with popular AI frameworks and agent builders, making it easy to add semantic search and context retrieval to your existing applications.

Available Integrations

LlamaIndex

Python tool for LlamaIndex agents

Vercel AI SDK

TypeScript integration for Vercel AI SDK

Pipedream

No-code workflows and agent builder

Google Antigravity

Google’s agent framework integration

Quick Comparison

Framework	Language	Use Case	Installation
LlamaIndex	Python	AI agents with tool calling	pip install llama-index-tools-airweave
Vercel AI SDK	TypeScript	AI applications with streaming	npm install @airweave/vercel-ai-sdk
Pipedream	No-code/JavaScript	Workflows and automations	Connect via Pipedream UI
Google Antigravity	Python	Google’s agent framework	Built-in integration

LlamaIndex

The llama-index-tools-airweave package provides an AirweaveToolSpec for LlamaIndex agents.

Installation

pip install llama-index llama-index-tools-airweave

Quick Start

import os
import asyncio
from llama_index.tools.airweave import AirweaveToolSpec
from llama_index.core.agent.workflow import FunctionAgent
from llama_index.llms.openai import OpenAI

# Initialize the Airweave tool
airweave_tool = AirweaveToolSpec(
    api_key=os.environ["AIRWEAVE_API_KEY"],
)

# Create an agent with the Airweave tools
agent = FunctionAgent(
    tools=airweave_tool.to_tool_list(),
    llm=OpenAI(model="gpt-4o-mini"),
    system_prompt="""You are a helpful assistant that can search through
    Airweave collections to answer questions about your organization's data.""",
)

# Use the agent
async def main():
    response = await agent.run(
        "Search the finance-data collection for Q4 revenue reports"
    )
    print(response)

if __name__ == "__main__":
    asyncio.run(main())

Available Tools

• search_collection: Simple search with default settings
• advanced_search_collection: Full control over retrieval parameters
• search_and_generate_answer: RAG-style direct answers
• list_collections: List all available collections
• get_collection_info: Get collection details

See the example code above for LlamaIndex integration.

Vercel AI SDK

The @airweave/vercel-ai-sdk package provides an airweaveSearch tool for the Vercel AI SDK.

Installation

npm install ai @ai-sdk/openai @airweave/vercel-ai-sdk

Quick Start

import { generateText } from 'ai';
import { openai } from '@ai-sdk/openai';
import { airweaveSearch } from '@airweave/vercel-ai-sdk';

const { text } = await generateText({
  model: openai('gpt-4o'),
  prompt: 'What were the key decisions from last week?',
  tools: {
    search: airweaveSearch({
      defaultCollection: 'my-knowledge-base',
    }),
  },
  maxSteps: 3,
});

console.log(text);

Configuration Options

airweaveSearch({
  apiKey: 'your-api-key',              // Defaults to AIRWEAVE_API_KEY env
  defaultCollection: 'my-collection',  // Default collection to search
  defaultLimit: 20,                    // Max results per search (default: 10)
  generateAnswer: true,                // Generate AI answer (default: false)
  expandQuery: true,                   // Query expansion (default: true)
  rerank: true,                        // Rerank results (default: true)
  baseUrl: 'https://custom.url',       // For self-hosted instances
});

See the example code above for Vercel AI SDK integration.

Pipedream

Pipedream is a low-code platform for building workflows and agents. The Airweave integration provides actions for searching collections, managing data, and triggering syncs.

Setup

1. Create Workflow

   Open Pipedream and create a new workflow

2. Add Airweave Action

   Add an Airweave action step to your workflow

3. Connect Account

   Click Connect Account and enter your Airweave API key

4. Configure Action

   Select an action (Search Collection, List Collections, etc.)

Available Actions

• Search Collection: Semantic and keyword search
• List Collections: Get all collections
• Get Collection: Collection details
• Create Collection: Create new collection
• Delete Collection: Remove collection
• List Sources: Available data connectors
• Trigger Source Connection Sync: Manual data sync

Example: Slack Q&A Bot

1. Trigger: Slack - New Slash Command
2. Action: Airweave - Search Collection (with responseType: "completion")
3. Action: Slack - Reply to Command

See the example workflow above for Pipedream integration.

Google Antigravity

Google Antigravity is Google’s agent framework. Airweave provides native integration for semantic search capabilities.

Setup

Airweave is available as a built-in integration in Google Antigravity. Configure it with your API key and collection ID.

from google.antigravity import Agent
from google.antigravity.integrations import AirweaveIntegration

# Initialize Airweave integration
airweave = AirweaveIntegration(
    api_key="your-api-key",
    collection_id="your-collection-id"
)

# Create agent with Airweave
agent = Agent(
    integrations=[airweave],
    instructions="You can search our knowledge base to answer questions."
)

# Use the agent
response = agent.run("What is our refund policy?")
print(response)

See the example code above for Google Antigravity integration.

Other Integrations

LangChain

Use Airweave with LangChain via the REST API or SDKs:

from langchain.tools import Tool
from airweave import AirweaveSDK

client = AirweaveSDK(api_key="your-key")

def airweave_search(query: str) -> str:
    results = client.collections.search(
        readable_id="my-collection",
        query=query,
        response_type="completion"
    )
    return results.completion

tool = Tool(
    name="AirweaveSearch",
    func=airweave_search,
    description="Search the knowledge base for relevant information"
)

CrewAI

Integrate Airweave with CrewAI agents:

from crewai import Agent, Task, Crew
from airweave import AirweaveSDK

client = AirweaveSDK(api_key="your-key")

def search_knowledge_base(query: str) -> str:
    results = client.collections.search(
        readable_id="company-docs",
        query=query
    )
    return "\n".join([r["payload"]["md_content"] for r in results.results[:3]])

researcher = Agent(
    role="Knowledge Base Researcher",
    goal="Find relevant information from the knowledge base",
    tools=[search_knowledge_base]
)

Custom Integration

For frameworks not listed above, use the Python SDK or TypeScript SDK to build custom integrations:

Python

from airweave import AirweaveSDK

client = AirweaveSDK(api_key="your-key")

# Your custom integration logic
results = client.collections.search(
    readable_id="my-collection",
    query="user query"
)

Common Patterns

RAG (Retrieval-Augmented Generation)

All integrations support RAG patterns by setting response_type="completion" or generateAnswer: true:

LlamaIndex

result = airweave_tool.search_and_generate_answer(
    collection_id="docs",
    query="What is our refund policy?",
    use_reranking=True
)
print(result)  # Direct answer

Advanced Search Parameters

All integrations support advanced search parameters:

• Recency Bias: Weight recent content higher
• Score Threshold: Filter low-relevance results
• Search Method: Choose hybrid, neural, or keyword search
• Reranking: LLM-based result reranking
• Query Expansion: Automatic query variations

LlamaIndex

result = airweave_tool.advanced_search_collection(
    collection_id="news",
    query="AI breakthroughs",
    temporal_relevance=0.8,  # Favor recent content
    retrieval_strategy="hybrid",
    rerank=True
)

Need Help?

SDK Documentation

View Python and TypeScript SDK docs

API Reference

Complete API documentation

Discord Community

Get help from the community

GitHub

View source code and examples