--:--:--LOCAL ·--:--UTC
    MKT CAP$2.85T+0.4%
    24H VOL$2082.8B
    BTC DOM59.1%
    HEALTH62NEUTRAL
    SHORT-TERM48BEARISH
    LONG-TERM77BULLISH
    OI$14.0B
    24H LIQ$41M
    LONG/SHORT55.7% / 44.3%
    REGIME (LT)DISTRIBUTION / POST-PEAK DELEVERAGING
    REGIME (ST)RANGE / LOW VOLATILITY
    HL OI$11.7B
    WHALESSHORT 45.4%
    MKT CAP$2.85T+0.4%
    24H VOL$2082.8B
    BTC DOM59.1%
    HEALTH62NEUTRAL
    SHORT-TERM48BEARISH
    LONG-TERM77BULLISH
    OI$14.0B
    24H LIQ$41M
    LONG/SHORT55.7% / 44.3%
    REGIME (LT)DISTRIBUTION / POST-PEAK DELEVERAGING
    REGIME (ST)RANGE / LOW VOLATILITY
    HL OI$11.7B
    WHALESSHORT 45.4%
    Plug real-time crypto data into your AI agent — one command: Get your free API key →
    LoginGet API Key
    Coming soon · Waitlist open

    AI Agent
    Trading Academy

    Learn to build AI agents that read live crypto markets, test their own ideas and paper-trade while you sleep — from your very first MCP call to an autonomous research desk.

    Free to join. You'll also get the CryptoDataAPI newsletter — unsubscribe anytime.

    • 6 levels
    • 190+ data endpoints
    • 50 strategies
    • 24/7 markets
    The curriculum

    Six levels, from first prompt
    to autonomous desk

    Every level ends with something real that runs. No theory marathons: you build an agent on day one and level it up all the way through.

    01 Level 1 · Foundations

    Wire up your first agent

    Connect Claude, ChatGPT or Grok to live market data over MCP and ask your first real questions about the market.

    You'll shipAn agent that briefs you on the market every morning
    02 Level 2 · Market literacy

    Read the market like a quant

    Funding, open interest, liquidations, regimes and health scores, and how an agent should weigh each one.

    You'll shipA market-regime dashboard your agent explains
    03 Level 3 · Strategy design

    Turn ideas into hypotheses

    Mine AlgoBrain's 4,900-node strategy wiki and turn vague hunches into rules an agent can actually test.

    You'll shipThree testable strategy specs
    04 Level 4 · Backtesting

    Backtest without lying to yourself

    Look-ahead bias, survivorship, fees and funding, and walk-forward tests on real Hyperliquid history.

    You'll shipA backtest report you'd defend in public
    05 Level 5 · Automation

    Run a paper-trading desk

    Schedule your agent, give it memory, journal every decision and fence it in with hard risk rules.

    You'll shipA 24/7 paper desk with a decision journal
    06 Level 6 · Graduation

    Prove your agent has skill

    Score your agent against a coin flip, monitor it for drift and decide, with evidence, whether it has an edge.

    You'll shipA public scorecard for your agent
    What's inside

    Everything you need to build, test and ship

    Build-along video lessons

    Short, dense lessons where we build each agent live on screen, then you rebuild it yourself with live data.

    Agent playground

    A pre-wired sandbox with live market data, prompts and notebooks ready to go. No setup and no API wrangling.

    Prompt & skill packs

    Battle-tested system prompts, agent skills and checklists for every level, ready to drop into your own agent.

    Agent leaderboard

    Pit your paper-trading agent against other students' agents. Every one is scored against a baseline, with no cherry-picking.

    Live build sessions

    Regular live sessions where we take on a market event as it happens and build the agent that would have caught it.

    Agent Builder certificate

    Graduate with a verifiable certificate and a public scorecard showing your agent's track record.

    Real tools, not toys

    Learn on the stack you'll actually use

    The Academy is built on the same production data your agents will run on after you graduate. Here's some of what's already live:

    agent_loop.pylevel 05
    # Level 5: a paper-trading loop your agent runs 24/7
    while desk.is_open():
        market = cda.snapshot("BTC", "ETH", "SOL")
        regime = cda.regime()            # trend / range / squeeze
    
        idea = agent.think(market, regime,
                           memory=journal.recent(20))
    
        if risk.approves(idea, max_risk=0.005):
            desk.paper_trade(idea)
    
        journal.log(idea, why=agent.reasoning)
        desk.sleep(minutes=15)
    Who it's for

    Built for three kinds of builders

    Traders

    You know the markets but are tired of staring at charts. Get an AI co-pilot that watches everything and explains what it sees.

    Developers

    You can ship code but trading is a black box. Learn the market structure that separates real signals from noise.

    Agent builders

    You're building with LLMs and want a real-world domain. Crypto runs 24/7, its data is open, and it gives instant feedback.

    The road to launch

    Get on the list, get in first

    01
    Now

    Waitlist open

    Register your interest with just an email. Early sign-ups get first word on everything we build.

    03
    Then

    Doors open

    The full Academy opens to everyone, with all six levels, the playground and the leaderboard.

    FAQ

    Questions, answered

    When does the Academy launch?

    It's in development now. A founding cohort from the waitlist gets in first, then it opens to everyone. Register your interest and you'll hear the moment doors open.

    Do I need to know how to code?

    No. The early levels work entirely through AI assistants like Claude, ChatGPT and Grok connected to live market data. Later levels bring in Python for backtesting and automation, taught from zero.

    How much will it cost?

    Pricing will be announced at launch. Joining the waitlist is free, and waitlist members hear about founding-member perks first.

    What data will my agents use?

    CryptoDataAPI's 190+ endpoints of live and historical market data: prices, funding, open interest, liquidations, order books, regimes, sentiment and full backtesting archives.

    Is this financial advice?

    No. The Academy is educational. It teaches you how to build, test and evaluate AI trading agents, and every agent starts on paper. Nothing in it is a recommendation to buy or sell anything.

    Founding cohort · limited spots

    Be first through the door

    Register your interest and we'll email you the moment the Academy opens, before anyone else.

    Free to join. You'll also get the CryptoDataAPI newsletter — unsubscribe anytime.

    • First access to the founding cohort
    • Founding-member perks
    • Shape what we teach

    The AI Agent Trading Academy is educational content and is not financial advice. Trading crypto assets carries a high risk of loss. Agents in the Academy are built and evaluated on paper first. Curriculum and features may change before launch.