80+ integrations. Every tool your agents need.
Pre-built connectors for the LLMs, databases, vector stores, and SaaS tools your agents rely on. Extend with your own via the Tool SDK in one decorator.
Integration categories
LLM Providers
OpenAI, Anthropic, Google Gemini, Mistral, Cohere, Together AI, Ollama (local). Bring your own API keys. Diaflow doesn't mark up LLM costs.
Vector Stores
Pinecone, Weaviate, Qdrant, Chroma, pgvector, Milvus. Semantic search and hybrid retrieval backed by any of these stores, unified behind the Memory API.
Databases
PostgreSQL, MySQL, MongoDB, Redis, Elasticsearch, BigQuery, Snowflake. Query, read, and write as tool calls — schema-inferred automatically.
SaaS & APIs
Jira, Linear, Slack, Notion, Salesforce, HubSpot, Zendesk, GitHub, PagerDuty. Each with pre-built auth handling and schema-typed action sets.
Tool SDK
Write any function with the @tool decorator and it becomes an agent tool with auto-generated schema. Works with any Python callable.
Webhooks & Events
Trigger agent runs from external events. Send results back to webhooks. Integrate with Zapier, Make, n8n, or your own event bus.
Write your own tool in one decorator
No boilerplate. No schema files. Just decorate a function and it's available to any agent on your account.
from diaflow import tool
@tool(description="Fetch the latest price for a stock ticker")
def get_stock_price(ticker: str) -> dict:
"""Returns price, volume, and change."""
return my_market_api.fetch(ticker)
# Now any agent can call it:
agent.run(
"What is AAPL trading at today?",
tools=[get_stock_price]
)Connect your first tool today
Free tier includes full integration access.