Bybit Savings review: a 12% rate with a 500 USDT cap

Bybit is one of the largest derivatives-first exchanges in crypto, and its savings product sits beside the trading engine rather than in front of it. Two shapes are offered: Flexible, which can be entered and exited at any time at a floating rate, and Fixed Term, which locks a balance for a set period at a fixed APR with a fixed yield amount stated up front.

The reason Bybit appears near the top of stablecoin rate tables is a single published number: 12.00% APR on flexible USDT. The reason it should not be read that way is the three-line tier table printed directly beneath it. Understanding that table takes about thirty seconds and changes the answer completely.

There is a second problem with reviewing this platform, and we are going to be explicit about it throughout. When we attempted to verify the product from primary sources on 16 September 2026, Bybit's help-centre articles returned "not supported on this site" and the Easy Earn terms and conditions page was marked noindex and returned no body. Everything mechanical below therefore comes from Bybit Learn, which is operated by Bybit but is marketing-tier documentation rather than contractual terms.

Which engine is paying you

This is the first question we ask of any earn product, and on Bybit it does not have a confirmed answer. A savings product on a derivatives exchange has an obvious candidate mechanism — lending deposited assets to margin traders, who pay to borrow so they can go long or short with leverage. That is exactly what OKX states plainly about its own Simple Earn product. Bybit does not state it. It does not state anything else either.

We are not going to assert a mechanism the provider has not disclosed. What we can say is that the shape of the tier table is consistent with a promotional rate on a small first tranche sitting on top of a much lower base rate, and a 0.28% base is well below what genuine lending demand pays. For comparison, supplying USDT to Aave v3 on Ethereum returned 3.11% on 16 September 2026 at around 86% utilisation. A 0.28% rate is not a lending rate; it is closer to nothing. The 12.00% top tier, meanwhile, is far above any sustainable lending rate and behaves like customer acquisition spend. Ourstablecoin yield page sets out the four mechanisms that can genuinely produce a dollar return.

The tier table, and the arithmetic behind it

Bybit publishes three tiers on flexible USDT. The table below applies them to a 10,000 USDT balance on the marginal reading — each slice priced at its own tier, which is how tiered savings products normally work.

Both readings lead to the same editorial conclusion. The advertised rate is available only on an amount smaller than most people's first deposit, and a serious stablecoin allocation earns something between a quarter and nine tenths of one per cent. Against a three-month US Treasury bill at 3.97% on 15 September 2026, that is a negative real proposition before considering any counterparty risk at all.

Promotional layers on top

Bybit also runs APR Boosters — vouchers earned through the Rewards Hub, one example being a 10% booster, which are stackable and described as potentially raising rates to 60%. VIP members receive separate, higher-APR staking offers. Every one of these is promotional and time-limited by construction. A 60% APR is not a yield; it is a marketing budget divided by a quota. Treating boosted rates as a planning assumption is the specific error that makes exchange earn products look better than they are, and the same pattern appears in ourearn free crypto coverage.

Fees, lock-up and getting your money back

Flexible positions can be unstaked at any time. Fixed Term positions cannot be withdrawn during the lock at all — the APR and the yield amount are fixed at entry, and so is the commitment. Minimums are the detail worth reading twice, because they move by a factor of five hundred across two products on the same asset: flexible USDT starts at 1 USDT, while the one-day USDT fixed term requires 500 USDT.

The T+1 accrual and T+2 distribution schedule means a deposit does not begin earning on the day it lands, and does not pay until two days later. A balance parked for forty-eight hours between trades earns nothing. Rewards land in the Funding Account rather than the trading account, so they do not automatically compound unless Auto-Savings is switched on.

Availability, KYC and who is excluded

KYC is mandatory and runs in three tiers. Standard requires a government ID and a face scan and carries a 1 million USDT daily withdrawal limit. Advanced adds proof of address and raises that to 2 million USDT. Pro is institutional enhanced due diligence. Identity documents issued by an excluded jurisdiction block onboarding outright, which is a stricter test than a residence check.

Roughly fourteen jurisdictions are restricted: the United States, Canada, Chinese Mainland, Hong Kong, Singapore, Uzbekistan, Iran, Syria, Sudan, North Korea, Cuba, Dubai on the global platform, and Crimea, Donetsk, Luhansk and Sevastopol. Two markets moved back the other way: the United Kingdom in December 2025, through a partnership with FCA-authorised Archax covering spot trading only and no derivatives, and India in September 2025. EEA customers are served through Bybit EU rather than the global platform. Ouravailability tracker keeps the current position for each venue.

Licensing and what each authorisation actually covers

Bybit EU GmbH holds a MiCA CASP authorisation granted by Austria's Financial Market Authority on 28 May 2025. The group also holds a UAE SCA Virtual Asset Platform Operator licence from October 2025, an FIU registration in India, an Astana Financial Services Authority registration in Kazakhstan and a National Bank of Georgia VASP registration.

Each of these regulates the provision of crypto-asset services: conduct, governance, disclosure, complaints and anti-money-laundering obligations. None of them is a deposit guarantee, and none of them makes a regulator responsible for the value of a savings balance. The UK arrangement is narrower still — the Archax partnership is FCA-authorised for spot activity, and an FCA authorisation held by a partner firm does not extend to a savings product offered by the offshore entity.

Custody, the 2025 exploit and proof of reserves

In February 2025 Bybit suffered a cold-wallet exploit of roughly US$1.4 billion, one of the largest single custody losses in the industry's history. The incident sat outside the savings product and Bybit continued operating, which is itself a data point about the exchange's balance sheet depth. But a savings balance is held in exchange custody like any other balance, so the event is material context rather than trivia.

Bybit publishes proof of reserves on a Merkle-tree basis. We did not verify the current attestation date in this pass and will not characterise it. What we can say is that a Merkle-tree proof demonstrates that customer balances are matched by on-chain holdings at a point in time. It says nothing about liabilities that are not customer balances, and nothing about whether the assets are encumbered.

Risks specific to this platform

The first risk is documentary. On most platforms we can read the terms, identify the contracting entity and establish whether a savings balance is a loan to the company or a custodial holding. On Bybit we could not, because the pages that carry that information would not load. That is not a hypothetical concern: the distinction between a custodial balance and an unsecured loan is exactly what determined outcomes in theCeFi failures of 2022, and a reader who cannot establish which one they hold cannot price the risk at all.

The second is presentation. A 12% headline attached to a 500 USDT cap will be repeated by aggregators and comparison sites without the cap, and it has been. The information is published, so this is not concealment. It is a structure that produces misleading third-party coverage as a predictable by-product, and Bybit benefits from that coverage.

The third is custody concentration. Using an exchange savings product means your assets sit in the same custody perimeter as the trading business. The February 2025 exploit demonstrated what that perimeter is worth when it fails, and no savings rate on this platform compensates for holding a large balance there.

How it compares to two alternatives on this site

Against OKX, the contrast is about disclosure rather than rate. OKX states in its own rules document that Simple Earn assets are pooled and loaned to borrowers including margin traders, publishes a 15% commission, and describes hourly distribution and redemption mechanics. Bybit publishes a higher headline number and almost none of the surrounding facts. If you want to know what you are buying before you buy it, OKX is the better-documented of the two by a wide margin, even though what it documents is a lending product with real credit exposure.

Against KuCoin, the two are more alike. Both publish wide rate ranges whose top ends are promotional and quota-limited, both leave the fee undisclosed, and both are offshore exchanges serving EEA customers through a recently authorised European subsidiary. KuCoin discloses more product variety and a clearer set of structured products; Bybit discloses a clearer tier table. Neither tells you where the money comes from. For readers who want that answer, the useful comparison is witha lender that names its loan book.

The verdict

Bybit's savings product is a small, capped promotional rate sitting on top of a base rate close to zero, wrapped in documentation we could not read. The tier table deserves credit for existing — plenty of venues do not publish one — but the product it describes is not a competitive place to hold stablecoins once your balance passes 1,000 USDT, and at 0.28% it is not competitive against anything, including leaving the money in a money-market fund.

The absence of readable terms is the finding that should carry the most weight. We can tell you what Bybit's marketing documentation says about accrual, minimums and tiers. We cannot tell you what claim you hold, who you hold it against, or what happens to it if the entity fails, because the pages that would answer those questions returned nothing on 16 September 2026. That is a gap worth weighing more heavily than any rate. See how the rest of the market handles the same questions on ourearn platform comparison.