How to use your AI Subaccount with Double-Win on Bybit

Double-Win is a Bybit structured product that’s built around one question: Will an asset's price stay inside a defined range, or break out of it by a set expiration time? Pairing it with an AI Subaccount lets you delegate the timing and monitoring of that question to an AI assistant, within strict boundaries you define, rather than tracking range breakouts yourself.

This guide describes how Double-Win actually settles, including the total-loss scenario that makes it different from most yield products, and how an AI Subaccount can manage Double-Win positions on your behalf.

Key Takeaways:

• Double-Win pays out only if the settlement price lands outside your chosen range. If it settles inside the range, you lose the entire investment amount — a stated, by-design risk, not an edge case.

• An AI Subaccount lets you delegate Double-Win decisions to an AI assistant within asset caps, transfer controls and leverage limits you define, rather than enabling unrestricted account access.

• Because Double-Win orders cannot be modified or canceled once subscribed, the AI assistant's instructions need to be precise before any order is placed.

What Double-Win does

Double-Win comes in two modes: Classic and Custom. Classic mode offers prebuilt plans listed by coin, estimated leverage, Double-Win range and expiration. Custom mode lets you select the coin and duration, enter an investment amount and adjust the range yourself, using a result simulator to estimate profit and loss before confirming.

Every plan settles around the same two-outcome structure.

This is why Double-Win is not a directional long or short position. Instead, you’re betting on whether the price stays contained or breaks out of a range you’ve set. Before confirming your order, Bybit requires acknowledgment that the order cannot be modified or canceled, and that a settlement price inside the range results in total loss of the investment.

Why pair Double-Win with an AI Subaccount?

Double-Win plans run on short, fixed tenors, and favorable range and leverage combinations shift as volatility changes. An AI Subaccount lets an AI assistant monitor available plans and act on instructions you set in advance, without access to your Main Account balance. 

Every permission — asset cap, transfer controls, leverage limits — is something you control directly. The AI assistant can only commit the funds and leverage you have explicitly allowed.

Whom this product is for

• Users comfortable with a total-loss outcome on individual positions in exchange for a defined payoff structure.

• Traders who want range-based exposure without manually tracking plan expirations and range parameters.

• Holders of idle stablecoins or crypto assets looking for a structured, time-boxed way to deploy them.

Setup guide

This guide assumes that you’ve already created an AI Subaccount, and have connected it to an AI assistant.

1. Fund your AI Subaccount: Transfer USDT or the relevant coin in, and set an asset cap that reflects how much you’re willing to risk on a total-loss basis.

2. Set appropriate leverage limits: Maximum margin and contract leverage are uncapped by default. Cap them deliberately, given Double-Win's leveraged payoff structure.

3. Give your AI assistant a standing instruction: Specify the coin and preferred tenor, and specify how the Double-Win range should be set relative to the current price.

4. Choose Classic or Custom mode: Tell your AI assistant whether to select an existing Classic plan or configure a Custom range and duration.

5. Confirm before placing any order: Since orders cannot be modified or canceled after subscription, be sure to carefully review the proposed parameters first.

6. Monitor positions and settlement: Track active orders and outcomes from the Earn Account or Subaccount order history.

Use cases

• Deploy idle USDT into a short-tenor plan with a wide range, accepting a lower payout for a higher probability the price stays outside the range.

• Use Custom mode to set a narrower range around a price level you believe is unlikely to hold, in exchange for higher estimated leverage and payout.

• Run several smaller positions across different coins or tenors, rather than concentrating risk in one plan.

Best practices

• Never allocate more than you can lose entirely. A settlement price inside the range means zero return of principal, not a partial loss.

• Use the result simulator before confirming your order. Estimate profit and loss using your investment amount, range and an anticipated price scenario.

• Keep position sizes proportional. Spread allocations across multiple plans, rather than concentrating capital in one order.

• Review your AI assistant's range and tenor choices regularly. Volatility changes which ranges are reasonable, so static instructions can go stale.

Risk considerations

• Total loss, one possibility, is a real, designed outcome. A settlement price within the range means the entire investment is lost — one of two possible outcomes by design, not a tail risk.

• Orders cannot be modified or canceled. Once you’ve subscribed, a position is locked until expiration or early redemption.

• Early redemption has its own trade-offs. You can redeem up to one hour before expiration, but proceeds will not follow the payoff formula at expiration, and the amount may fall short of the estimate. With 90% slippage protection enabled by default, redemption fails if the amount falls below 90% of the estimate, leaving your plan unaffected.

• All trading risk sits with you. Bybit does not control, monitor or guarantee an AI assistant's outcomes. Please review permissions carefully before funding your Subaccount.

The bottom line

Double-Win pays based on a range, not a direction, and the cost of being wrong is the entire position. An AI Subaccount can streamline the way that positions get monitored and placed, but it doesn’t change the underlying risk: every order carries a real chance of total loss, and that should shape position sizing and leverage more than anything else in this guide.

#LearnWithBybit