Hyperliquid testnet: paper trade perps before you risk USDC

What the Hyperliquid testnet is

Hyperliquid runs a full copy of its exchange on a separate testnet, with its own API (api.hyperliquid-testnet.xyz), its own order books and test USDC that has no value. Everything that matters for automation works the same way there: agent approvals, order types, margin, leverage and funding. That makes it the right place to prove an automated setup before it touches real money.

TensorTrader treats the Hyperliquid testnet as a confirmed paper venue. Balances come from the public clearinghouse state for perps and spot, and every new Hyperliquid connection defaults to the testnet card. The engine also never silently flips a failing mainnet key to testnet or the reverse, so the environment you chose is always the one you trade.

Funding a testnet account

Testnet USDC comes from Hyperliquid's testnet faucet. Eligibility rules for the faucet are set by Hyperliquid and change from time to time, so check their current documentation before you start. Once funded, the TensorTrader connect panel shows the testnet balance on its network card; if both testnet and mainnet read empty, the panel links you to the Hyperliquid app to fund an account.

Hyperliquid accounts use a unified model where spot USDC can count as perp margin. TensorTrader reads equity as perp account value plus spot USDC, and before trading it moves spot USDC into perps when your account needs that step. In Unified Account mode, the default for new accounts, it skips the transfer because spot USDC already counts.

Connect the testnet to TensorTrader

1. 1Open Settings, Exchange Auth, and pick Hyperliquid.
2. 2Press Connect Wallet and choose the "Testnet – Paper / test" card.
3. 3Approve the TensorTrader agent in MetaMask. The wallet switches to Arbitrum Sepolia for the signature.
4. 4Wait for the green status and the testnet balance.
5. 5Build a small set of TradingView alerts with the extension’s Batch Create tab, pointed at this Hyperliquid key.
6. 6Watch the first alerts fire, fill and close on the dashboard before adding more.

What to test on testnet

alert delivery

Does every TradingView alert reach TensorTrader, and does each one map to the right symbol and direction?

sizing

Do position sizes and leverage match your policy after the venue clamp?

exits

Do stop-loss and take-profit orders appear right after the fill, and do they move when the engine ratchets?

closes

Do closed trades book profit and fees from Hyperliquid fills, and does the dashboard match the exchange?

failure paths

What happens when you pause trading, disconnect the key, or send an alert for an unsupported symbol?

What testnet cannot tell you

Testnet order books are thinner and behave differently from mainnet. Fills can be easier or harder than they would be with real liquidity, and prices can drift from mainnet. Use testnet to prove mechanics (delivery, sizing, exits, accounting), not to estimate profitability. For profitability, look at forward results on paper over many trades, count fees, and assume live results will be somewhat worse.

Our own Hyperliquid testnet book is a useful warning: across 105 closed TT-Autotune trades between 6 and 19 August 2026, it lost 11.9% of its starting testnet equity, and fees made up a meaningful part of the loss. Testnet exists precisely so that lessons like this cost nothing.

A two-week testnet plan

Write down what you expect before each step, then compare. The point of testnet is to be surprised cheaply: an alert that maps to the wrong symbol, a size that is larger than you meant, a stop that sits further away than you pictured. Every surprise you find here is one you do not pay for on mainnet.

1. 1Days 1-2: connect, fire one alert by hand from TradingView and confirm the fill and exits.
2. 2Days 3-7: run a small batch (three liquid tokens, one timeframe) and check every close against the exchange.
3. 3Days 8-14: add a second timeframe as a DCA leg and compare net ROI after fees per timeframe.
4. 4Day 14: decide. Keep what paid for its fees, cut what did not, and only then consider a mainnet agent.

Testnet vs paper simulation

Some platforms "paper trade" by simulating fills from a price feed. Testnet is different: orders really go to an exchange matching engine, with real order types, rejections and partial fills, just with test money. That is why TensorTrader prefers venue testnets and demo environments over internal simulation for every venue that offers one, and fails closed on venues that have none.

Frequently asked questions

Is Hyperliquid testnet free?

Yes. Test USDC has no value; you only need a browser wallet to sign the agent approval.

Does TensorTrader default to testnet?

Yes. Every new Hyperliquid connection starts on the testnet card, and mainnet is an explicit choice.

Can I switch from testnet to mainnet later?

Yes. Connect again, choose the mainnet card and approve a mainnet agent. Choosing mainnet is the live opt-in.

Keep reading

• Hyperliquid trading bot: how to automate Hyperliquid perps safely

  The complete guide to automating Hyperliquid: agent wallets, testnet, TradingView and GoCharting alerts, leverage and margin, fees and funding, portfolio mode, and the real results of our own testnet books.

• What is a Hyperliquid API (agent) wallet, and is it safe?

  A Hyperliquid API wallet is a separate key your main wallet authorizes to trade but never withdraw. Here is how it works, what it can and cannot do, and how TensorTrader creates one without ever seeing your seed phrase.

• Paper trading and testnets on every exchange TensorTrader supports

  Which crypto exchanges and brokers offer a testnet, demo or paper environment, how TensorTrader uses each one, and which venues have no paper path at all.

• TradingView alerts to Hyperliquid: webhook setup, step by step

  Send TradingView indicator and strategy alerts to Hyperliquid perps through TensorTrader: connect an agent wallet, build alerts with the Chrome extension, and check the webhook message.