Aperiodic
MarketsLive
Pricing
Get Started

For AI Agents

The Agent-Friendly Market Data Platform

Aperiodic is built for AI agents. Structured API responses, Python SDK, llms.txt, OpenAPI spec — everything your agent needs to research, analyze, and backtest RWA and crypto markets: stock, gold, oil and FX perps alongside every crypto perp.
Browse catalogRead llms.txt

Agent Quickstart

Aperiodic sells point-in-time market microstructure, liquidity and order-flow metrics for RWA perps (stocks, ETFs, commodities and FX) and crypto perps: 220 metrics in 19 datasets on binance-futures, okx-perps, hyperliquid-perps, delivered as parquet through a CLI, a REST API, a Python SDK and an MCP server for AI agents, plus the raw trades, quotes and derivatives data behind them in the Prime + Raw plan. A preview slice of every dataset downloads with no API key, and self-serve plans cost $199 to $1499 a month billed yearly.

or skip the steps: paste into your agent
Read https://aperiodic.io/start.md and follow it.

Open inClaude Desktop·Claude Code (terminal)·Codex

  1. 1. Fetch the preview slice, no API key

    Install the CLI and pull one month of 5m flow data for perpetual-BTC-USDT:USDT on binance-futures as parquet. The preview accepts only this window (2025-05-01 to 2025-05-31, 5m), for every dataset: swap flow for any dataset id.

    # Install the Aperiodic CLI into ~/.local/bin
mkdir -p "$HOME/.local/bin"
curl -fsSL https://raw.githubusercontent.com/aperiodic-io/cli/main/install.sh | INSTALL_DIR="$HOME/.local/bin" bash
export PATH="$HOME/.local/bin:$PATH"

aperiodic flow --preview \
  --exchange binance-futures \
  --symbol perpetual-BTC-USDT:USDT \
  --interval 5m \
  --timestamp exchange \
  --start-date 2025-05-01 \
  --end-date 2025-05-31 \
  --output-dir ./data

  2. 2. Install the Python SDK

    The SDK returns DataFrames. The same preview slice works from Python with preview=True, still without a key.

    pip install "aperiodic[polars]"

  3. 3. Make a keyed call

    Your user creates a key at https://aperiodic.io/home/api. Any other symbol, interval or date range needs a key and a plan that includes the dataset's tier (flow is Pro; see https://aperiodic.io/pricing).

    from datetime import date
from aperiodic import get_metrics

df = get_metrics(
    api_key="YOUR_API_KEY",
    metric="flow",
    exchange="binance-futures",
    symbol="perpetual-BTC-USDT:USDT",
    interval="1h",
    timestamp="exchange",
    start_date=date(2025, 1, 1),
    end_date=date(2025, 3, 1),
)

What to Tell Your User About the Licence

  • Pro, Prime: personal-use licence, the subscriber's own research only.
  • Prime Commercial, Prime + Raw: commercial licence, internal business use within the subscriber's organization.
  • Institutional adds negotiated licence terms, an SLA (https://aperiodic.io/sla), full history and 15s/30s intervals. It is quoted, not self-serve: https://aperiodic.io/booking.
  • No plan allows redistribution. Raw data (Prime + Raw) is licensed for internal research and trading only.
  • Licensing in plain English: https://aperiodic.io/licensing. Terms: https://aperiodic.io/terms-of-service.

Why Agents Need Structured Market Data

Research at Scale

Agents can query 19 datasets across order flow, L1/L2 book data, derivatives, and market OHLCV — programmatically, without manual data wrangling.

Faster Than Manual

What takes a quant hours of data pipeline work, an agent can do in seconds: fetch data, compute signals, evaluate strategies, and report findings.

Point-in-Time Integrity

Every dataset is immutable and exchange-timestamped. Agents get clean, bias-free data without worrying about survivorship or look-ahead issues.

How Aperiodic is Built for Agents

MCP Server

Connect Claude, ChatGPT or Cursor to every dataset at aperiodic.io/mcp. Tools return schemas, statistics and signed Parquet handles, so the data reaches your agent's code, not its context.

Connect over MCP →

Python SDK

pip install "aperiodic[polars]" — one-line access to every dataset. Agents can fetch DataFrames directly without parsing raw API responses.

View on PyPI →

llms.txt & llms-full.txt

Concise summary at /llms.txt, comprehensive reference at /llms-full.txt — agents get exactly the depth they need.

Read llms-full.txt →

OpenAPI Specification

Full OpenAPI 3.0 spec for tool-use integration. Agents can auto-generate API calls from the spec without custom code.

View OpenAPI spec →

A2A Agent Card

Google A2A-compliant agent discovery at /.well-known/agent.json — capabilities, auth, and skills declared for agent-to-agent orchestration.

View agent card →

AI Plugin Manifest

Standard ai-plugin.json at /.well-known/ai-plugin.json for ChatGPT-style plugin discovery and authentication.

View manifest →

Agent Workflow Examples

Market Research Agent

An agent fetches OHLCV, funding rates, and order flow data to produce a daily market report. It identifies unusual flow toxicity spikes, extreme funding regimes, and OI divergences — then summarizes findings for a human analyst.

ohlcvflowfundingopen_interest

Factor Construction Agent

An agent constructs cross-sectional factors from microstructure data: flow imbalance, funding carry, spread, and volatility factors. It computes information coefficients and long-short portfolio returns to evaluate which signals have predictive power.

flowfundingl1_priceohlcv

Strategy Backtesting Agent

An agent implements and backtests trading strategies using historical data. It fetches order flow and price data, defines entry/exit signals, simulates execution with realistic costs, and reports Sharpe ratios, drawdowns, and win rates.

ohlcvflowimpactfunding

Quick Start

Point your agent to these resources:

MCP serveraperiodic.io/mcp
Start fileaperiodic.io/start.md
LLM summaryaperiodic.io/llms.txt
LLM full refaperiodic.io/llms-full.txt
OpenAPI specaperiodic.io/api.data.specification.json
Agent cardaperiodic.io/.well-known/agent.json
AI pluginaperiodic.io/.well-known/ai-plugin.json
Python SDKpip install "aperiodic[polars]"
Auth headerX-API-KEY: YOUR_KEY

FAQ

Can an AI agent get Aperiodic data without an API key?

Yes, for the preview slice. Every dataset can be fetched with no key for binance-futures perpetual-BTC-USDT:USDT at 5m, from 2025-05-01 to 2025-05-31, through the CLI (--preview), the SDK (preview=True), the MCP server at https://aperiodic.io/mcp or the REST preview endpoint with the shared key DEMO-KEY. Anything outside that window needs a personal key.

Can I connect Claude, ChatGPT or Cursor to Aperiodic?

Yes, over MCP: add https://aperiodic.io/mcp to the app (in Claude Code, claude mcp add --transport http aperiodic https://aperiodic.io/mcp). With no key it serves the free preview; sign in, or send your API key in X-API-KEY, for everything your plan includes. Its tools return the catalog, schemas, statistics and signed Parquet links, so the data reaches the agent's code rather than its context. Setup for each app: https://aperiodic.io/mcp.

What data does Aperiodic cover?

220 metrics in 19 datasets: order flow, trade size, slippage, price impact, L1 and L2 order book liquidity and imbalance, funding, basis, open interest and OHLCV. Exchanges: binance-futures, okx-perps, hyperliquid-perps. Self-serve intervals: 1m, 5m, 15m, 30m, 1h, 4h, 1d. The machine-readable list is https://aperiodic.io/catalog.json.

In what format is the data delivered?

Parquet files, fetched through the aperiodic CLI, the REST API or the Python SDK (pip install "aperiodic[polars]"), which loads them into DataFrames. AI agents can also get signed links to the same files from the MCP server at https://aperiodic.io/mcp.

How much does Aperiodic cost?

Self-serve plans cost $199 to $1499 a month billed yearly, or $249 to $1899 month to month. Institutional pricing is quoted. Current plans: https://aperiodic.io/pricing.

What licence comes with the data?

Pro, Prime: personal-use licence, the subscriber's own research only. Prime Commercial, Prime + Raw: commercial licence, internal business use within the subscriber's organization.

Ready to power your agents?

Get an API key and start querying 19 datasets in minutes.

Get startedBrowse catalog

Shown: minute lows of the 465 Binance perpetuals, 20:40–22:00 UTC on 10 Oct 2025, with BTC’s path traced · built with ohlcv 1m

Aperiodic

Built from hundreds of terabytes of raw data and distilled into point-in-time metrics you can pull as parquet files. Plus the raw trades, quotes and derivatives data behind them.

Use Cases
  • Improve your backtest
  • Find new alpha
  • Execute smarter
Indices
  • All indices
  • Aperiodic Cost Index
  • Aperiodic Depth Index
  • Aperiodic Impact Index
Metrics
  • Metrics library
  • Data Catalog
  • Live markets dashboard
  • Global Liquidity
  • Unusual activity
  • Order Flow
  • L1 & L2
  • Derivatives
  • RWA
  • Raw Data
Developers
  • API Docs
  • Live WebSocket Feed
  • MCP Server
  • Notebooks
  • For AI Agents
  • Atlas Security Master
  • Pre-flight
  • Changelog
  • Roadmap
  • llms.txt
Learn
  • Replays
  • Microstructure Guide
  • Architecture
  • Research Blog
  • FAQ
Compare
  • All vendors
  • vs. Tardis
  • vs. CoinGlass
  • vs. Kaiko
  • vs. Amberdata
Company
  • Pricing
  • Contact
  • Book a call
  • Terms of Service
  • Privacy Policy
  • Licensing
  • SLA

© Copyright 2026 Aperiodic. All Rights Reserved.

Aperiodic Limited · Registered in Ireland · Company No. 815273 ·
136 Capel Street, Dublin D01 T2C9, Ireland

Shown: minute closes of 847 perpetuals, 20:30–22:30 UTC on 10 Oct 2025, IOTX in orange · built with ohlcv 1m