The infrastructure layer for AI agents

Storage, communication, payments, and discovery in one protocol.

Get early access.

See How It Works
# Install
$ npm install @onchaindb/sdk
import { OnchainDB } from '@onchaindb/sdk';

const db = new OnchainDB({ apiKey: '...' });

// Store your agent's data
await db.store('memory', { user: 'alice', context: '...' });

// Query your data
const memories = await db.query('memory', { user: 'alice' });

// Query another agent's data (pay per query)
const prices = await db.query('flights@agent-b', { from: 'SFO' });
# Install
$ pip install onchaindb
from onchaindb import OnchainDB

db = OnchainDB(api_key='...')

# Store your agent's data
await db.store('memory', { 'user': 'alice', 'context': '...' })

# Query your data
memories = await db.query('memory', { 'user': 'alice' })

# Query another agent's data (pay per query)
prices = await db.query('flights@agent-b', { 'from': 'SFO' })

Manage your data

Monitor queries, track earnings, manage collections.

Works with your stack

Open-source data connectors

Connect your existing databases - no migration required

Celestia

Four primitives. One protocol.

Everything agents need to store, communicate, pay, and discover.

Storage

Persistent data layer for agents. Connect existing databases or use our managed storage. Immutability when you need it.

Communication

Standard protocol for agent-to-agent and agent-to-data communication. One API, universal access.

Payments

Native pay-per-query billing. Agents pay for data access automatically. Revenue flows to data providers.

Discovery

skills.md definitions let agents discover data sources, understand capabilities, and know costs upfront.

Build agents, not infrastructure

We handle storage, communication, payments, and discovery.

FeatureBuild Your OwnTraditional APIsOnchainDB
Pay-per-query billingBuild from scratchMonthly plans onlyBuilt-in
Automatic revenue sharingNot availableNot availableBuilt-in
No infrastructure to buildFull stack requiredCustom integrationSDK + skills.md
No data migrationETL pipelinesData export requiredConnect in place
Agent-optimized APIDesign yourselfHuman-first designNative support
Agent discovery (skills.md)Manual docsOpenAPI onlyAuto-generated
Immutable data optionComplex setupRarely availableOne flag

Track your earnings in real-time

Full visibility into queries, revenue, and data usage. Know exactly how your data is being consumed and how much you're earning - down to the individual query.

  • Query-level analytics
  • Revenue breakdown by collection
  • Usage trends and forecasting
Today's Revenue $1,247.50
12.4K Queries
$0.10 Avg/Query
1
Deploy connector onchaindb connect postgres://...
2
Set pricing $0.001 per query
3
Start earning Agents query, you get paid

Connect your data where it lives

Open-source data connectors plug into your existing infrastructure. No ETL pipelines, no data migration. Our universal query layer handles the rest.

  • PostgreSQL, ClickHouse, MongoDB, Supabase
  • Any REST or GraphQL API
  • Your data stays in your infrastructure

Questions

Any structured data that AI agents would find valuable: market data, knowledge bases, embeddings, research, proprietary datasets. If agents need it, you can monetize it.

No migration required. Our open-source data connectors plug directly into your existing infrastructure - ClickHouse, PostgreSQL, MongoDB, Supabase, or any API. Our universal query layer handles the rest, so your data stays where it is.

Data providers set prices per query. When an agent queries data, payment is initiated via x402. No invoices, no manual transfers. Top up with a card, cash out to your bank.

AI agents building applications on top of your data. As agent traffic grows 10-1000x vs human traffic, your query volume and revenue grow with it.

When someone queries your data, you earn. Revenue splits automatically between all data providers involved in the query. No negotiations, no manual tracking.

TypeScript SDK first. Python and other languages planned. Works with any agent framework - LangChain, AutoGPT, custom agents.

Every data source includes a skills.md file - a standardized definition that tells AI agents what data is available, how to query it, and what it costs. Agents can read and understand your data capabilities automatically.

We're building now. Sign up for early access to be first in line.

Ready to build on the agent infrastructure?

Join the protocol powering autonomous AI agents.

Data Providers

Publish your data and earn per query

Get Early Access

Developers

Build agents with access to rich data

Read the Docs