The Coins You Actually Can't Withdraw Right Now: A Live Cross-Exchange Tracker

Last updated: 30 June 2026

This article contains affiliate links. Yieldo may earn a commission at no extra cost to you.

TL;DR — 28% of Withdraw Routes and 56% of Deposit Routes Are Disabled Right Now

On any given day, ~28–30% of crypto withdrawal routes and ~55–60% of deposit routes are disabled across major centralized exchanges — this isn't a bug, it's a structural feature of how exchanges manage risk, but it's invisible until you try to move funds. On a snapshot from 25 June 2026, our cross-exchange tracker measured 28.9% of withdraw routes and 56.2% of deposit routes flagged as off. We collect this data every 30 minutes from the official APIs of seven supported CEXs (MEXC, Bybit, OKX, Bitget, Gate.io, KuCoin, and Binance as reference). Deposits get frozen roughly twice as often as withdrawals because the deposit pipeline has more fragile failure modes. Most restrictive in our snapshot: OKX (driven by 34 legacy OKTC routes) and Gate.io (broad catalog with long-tail freezes). Least restrictive: Bybit and Bitget. The full live dataset, including JSON and CSV feeds, lives at /fees/network-status under CC BY-NC 4.0. This article walks through the methodology, per-exchange profiles, why deposits asymmetrically lag, and how to check a network before you click withdraw.

The Hidden Default of Crypto You Don't See Until You Try to Withdraw

Crypto sells itself as the 24/7 alternative to legacy finance. No bankers' hours, no settlement windows, no "your money is in transit, check back Monday." The marketing copy is consistent across every major exchange home page: instant deposits, fast withdrawals, always on. Then you go to withdraw 200 USDT on TRC20 from one exchange to another, click the network dropdown, and discover that the network you needed is grayed out with a small "temporarily unavailable" tooltip. Or you copy a deposit address into your sending wallet and the receiving exchange has quietly disabled deposits on that route — meaning your funds are about to land in a chain endpoint nobody is currently crediting to your account.

This is the invisible default state of crypto in 2026. On any given day, somewhere between roughly one in three and one in four withdrawal routes (coin × network × exchange combinations) across the major CEXs are switched off. Deposits are worse: more than half are typically off. The numbers are not headline-worthy on their own — nobody announces "we're currently freezing 56% of our deposit endpoints" — but they reshape the practical experience of moving funds. The hero thesis of this study is simple: closed routes are not edge cases. They are the structural backdrop of cross-exchange transfers, and the tracker we publish at /fees/network-status is the first place where the aggregate picture becomes visible.

This piece is a research-study, not a how-to. If you are mid-transfer right now and your withdrawal is greyed out, the sister piece Withdrawal Network Disabled: What to Do covers the action steps. What follows here is the data behind the problem: how we measure it, what the per-exchange breakdown looks like, why deposits get frozen more often than withdrawals, and what the patterns reveal about how exchanges actually manage risk.

Methodology — How We Track withdraw_enabled and deposit_enabled Across the Supported Exchanges

If you came here from Hacker News, Reddit, or a search engine asking "is this data real," this section is for you. We make every methodology decision visible because the honest answer to "how many crypto routes are off right now" depends on how you count.

Data sources and API endpoints

Every supported exchange in our dataset is queried directly through its official REST endpoint. We do not scrape web UIs, do not buy third-party feeds, and do not rely on community-maintained lists. The endpoints we hit are the same ones the exchange's own wallet team uses to render the deposit and withdraw modal in their consumer apps:

• MEXC: api.mexc.com/api/v3/capital/config/getall (signed)
• Bybit: api.bybit.com/v5/asset/coin/query-info (signed)
• OKX: www.okx.com/api/v5/asset/currencies (signed)
• Bitget: api.bitget.com/api/v2/spot/public/coins (public)
• Gate.io: api.gateio.ws/api/v4/wallet/currency_chains (public)
• KuCoin: api.kucoin.com/api/v3/currencies (public)
• Binance: included as reference but not in our promoted route set

Each call returns a list of (coin × network) tuples with two boolean flags we care about: withdraw_enabled and deposit_enabled. We persist these into our withdrawal_fees table with a fetched_at timestamp and a 30-minute refresh cadence.

Sampling cadence (every 30 minutes since January 2026)

The current state table is refreshed on a 30-minute interval, driven by a Laravel scheduler that fans out parsing jobs across the queue. Continuous tracking began on 16 January 2026 with the creation of the withdrawal_fees table. A separate event log table, network_availability_events, came online on 25 June 2026; that table records each transition (withdraw_froze, withdraw_restored, deposit_froze, deposit_restored) with a recorded_at timestamp. This is the foundation for the v2 longitudinal study described at the end of this piece.

How we count "routes" (coin × network × exchange)

A "route" in our taxonomy is a single (exchange, coin, network) triple. USDT on TRC20 at MEXC is one route. USDT on TRC20 at Bybit is a different route. USDT on ERC20 at MEXC is yet another route. We do not collapse networks across exchanges, because the entire point of this study is that two exchanges can offer ostensibly the same network and one is open while the other is closed.

Across the seven supported exchanges, the total route population is in the low thousands — the exact number drifts daily as new coins list, old coins delist, and exchanges add or remove network options. For the 25 June 2026 snapshot the route count was high enough that a few-percentage-point shift represents tens of routes flipping in either direction.

What we measure vs. what we don't (snapshot v1 vs longitudinal v2)

We measure: the current state of withdraw_enabled and deposit_enabled on each route at the most recent fetch. We measure: which networks are most frequently flagged off in aggregate. We measure: per-exchange disabled-route shares.

We do not yet publish: per-route "frozen since" timestamps, mean time to recovery, freeze-frequency leaderboards, or classification of freezes by root cause (upgrade vs exploit vs liquidity). That is the v2 longitudinal study. The event log began accumulating on 25 June 2026 and needs several more weeks of data before publication.

Two important caveats:

• Stale guard. Our FeesAggregatorService::markStaleAsUnavailable() automatically flips withdraw_enabled to false for any route that hasn't been refreshed in two hours. These stale flips are written into the snapshot table but not into the event log — we keep "exchange said disabled" separate from "our data is stale." Within the snapshot itself, however, the two look identical, so a small fraction of the 28.9% / 56.2% headline numbers may reflect transient ingestion gaps rather than genuine exchange-side closures.
• Snapshot vs population. The numbers we report are point-in-time. They move daily. A single Ethereum mainnet upgrade event can swing the global percentage by five points within hours as exchanges precautionarily freeze ETH ERC20 routes. Use the ranges (~28–30%, ~55–60%) when interpreting trend, not point estimates.

Reproducibility — open data and live page

This is the part Show HN-style audiences usually skip to. All raw data is publicly available, with no login, under CC BY-NC 4.0:

• Live HTML page: /fees/network-status (updated every 30 min)
• Machine-readable JSON feed: /fees/network-status.json
• Full event history CSV dump: /fees/network-status.csv
• Embed widget: /embed/network-status (for blogs, dashboards, journalists)

Independent researchers, journalists, arbitrage developers, and data-curious readers are welcome to use any of this dataset under CC BY-NC 4.0 — please cite Yieldo as the source. Methodology details and provider list also live at About the data. As far as we have been able to verify, this is the first and only publicly accessible aggregated network-availability dataset in the industry. Individual exchanges publish per-coin freeze notices in their announcement centers, but no one else publishes the cross-CEX, time-stamped aggregate.

Snapshot Findings — What the Live Tracker Shows Today

The numbers below describe the state of the dataset on 25 June 2026 unless otherwise noted. For live current values, the widget below pulls fresh data from the same source. Treat any specific percentage with the understanding that it shifts on every 30-minute refresh.

Aggregate numbers across all supported exchanges

On 25 June 2026:

• 28.9% of withdrawal routes were flagged off across the seven supported exchanges
• 56.2% of deposit routes were flagged off across the same set
• Roughly 40–45% of routes were fully operational (both withdraw and deposit on)

Stated as ranges that hold across the recent weeks of data: roughly ~28–30% of withdraw routes and ~55–60% of deposit routes are disabled at any random sampling moment. The deposit number is the one that tends to surprise people. It is structurally roughly double the withdraw rate, and that gap is consistent across days.

The deposit/withdrawal asymmetry (deposits ~2× more often disabled)

The single most non-obvious finding in this study is the deposit-to-withdraw asymmetry. Naively, you might expect them to track each other closely — the same network being down should affect both directions. In practice, deposit endpoints are disabled roughly twice as often as withdraw endpoints. We dedicate a full section to this below. The short answer is that the deposit pipeline has more fragile failure modes (indexer lag, memo parsing, hot-wallet crediting) and exchanges are biased toward freezing inbound flow first because closing deposits doesn't make headlines — closing withdrawals does.

Most restrictive vs least restrictive exchanges (ranges)

A few caveats up front: catalog size matters. Gate.io and MEXC carry the broadest set of coins, so their absolute disabled counts are higher even when their per-route closure rates are similar to the average. Ranges below describe typical per-route disabled share, not absolute disabled count:

• OKX: ~30–40% withdraw disabled (driven heavily by the 34-route OKTC legacy block), ~55–65% deposit disabled
• Gate.io: ~30–40% withdraw disabled, ~60–70% deposit disabled (deepest long-tail catalog)
• MEXC: ~25–35% withdraw disabled, ~50–60% deposit disabled
• KuCoin: ~25–35% withdraw disabled, ~50–60% deposit disabled
• Bitget: ~25–30% withdraw disabled, ~50–60% deposit disabled
• Bybit: ~20–30% withdraw disabled, ~40–55% deposit disabled

So the simplest possible ranking, from most-restrictive to least-restrictive in our 25 June snapshot, runs OKX → Gate.io → MEXC → KuCoin → Bitget → Bybit. That ordering is not stable week to week — a single upgrade response can flip a position — but the spread (Bybit is consistently lower-restriction than OKX or Gate.io) is durable.

Most fragile vs most stable networks

By network, the pattern is roughly inverse to what marketing materials would suggest. The boring, "old" networks are the most reliably open. The new, marketed L2s see more closures, especially during upgrades and reorganizations.

• Most stable: TRC20 (USDT), ERC20 (when no Ethereum upgrade is in progress), BEP20, Polygon, Solana
• Most fragile in our window: newer L2s (Linea, Scroll, zkSync Era variants), legacy sunset chains (OKX OKTC, deprecated forks), and a long tail of low-volume Bridge networks
• Permanently off in the dataset: OKX OKTC (34 routes), MEXC ZKSYNC LITE for ETH, Bitget BITLAYER for BTC

The OKTC block deserves a closer look because it is the single most-cited outlier in cross-exchange fee aggregators. We'll cover it in the OKX profile below.

Per-Exchange Profiles — Closed Networks by Exchange

This section profiles each of the seven supported exchanges in our snapshot, with honest framing on why each exchange's closure profile looks the way it does. The CTA below each profile points to the exchange's page on our site and is included only because the network discussed in that section is typically open right there — this is a watchdog piece, not an affiliate roundup.

The widget below compares two of our most-discussed exchanges side by side. The cross-exchange routing case — same coin, two exchanges, one route open and one closed — is one of the most common practical reasons readers come to this dataset.

MEXC — closed networks profile and honest framing

MEXC carries one of the broadest coin catalogs in the industry, including the long tail of meme listings and low-cap altcoins. That breadth comes with a closure profile that skews toward narrow, low-volume cross-chain routes. The exchange aggressively pauses withdraw routes on niche altcoin networks after fast listings, especially when a token has just printed a sharp run-up and the desk anticipates arbitrage drain. Most of those temporary pauses lift within 24–72 hours. ZKSYNC LITE for ETH has been a long-standing exception: the zero-fee route has been disabled persistently in our data and is the example everybody points to when explaining the dataset.

A separate honest caveat: MEXC suspended the MX Holder withdrawal-discount program on 9 February 2026. That is not a network closure event — it is a fee schedule change — but it's the kind of MEXC-specific transition you'll want to know about if you're modeling cross-exchange transfer cost. Communication-wise, MEXC publishes maintenance notices in its Announcements Center but is sparser than peers on the company X account. The pattern recognition lesson: if you're moving a low-liquidity altcoin from MEXC, check the route status on /fees/network-status at the moment you actually click withdraw, not when you start planning.

On the routes that are typically open — USDT TRC20, USDT Polygon, popular stablecoin endpoints — MEXC remains one of the cheapest viable destinations. See the MEXC withdrawal fees overview for live per-network detail or jump straight to MEXC via https://yieldo.me/go/mexc?ctx=web_blog if you already know your route is open.

Bybit — closed networks profile and honest framing

Bybit consistently runs the lowest per-route disabled share in our snapshot. Its catalog is narrower than MEXC's or Gate's, which means fewer rare-altcoin routes that look closed because no one is moving size through them. The closure events you do see tend to be scheduled-maintenance windows announced in advance via the Bybit X account or the Help Center.

The visible exception was the 21 February 2025 Lazarus Group hack, in which ~$1.5B in ETH was drained from a Bybit cold wallet. Bybit froze a broad set of ETH-network withdraw routes within the first 24 hours and reopened them progressively over the next 24–72 hours; the losses were covered in full from insurance reserves within 72 hours. As a closure-response case study, this is the most thoroughly documented "sharp and fast" recovery in the dataset and is referenced repeatedly in the literature. It is also why Bybit's post-2025 closure pattern includes more precautionary maintenance windows than its pre-2025 baseline — the security perimeter tightened, and that perimeter manifests as more proactive route-pausing during high-volatility on-chain events.

Bybit also offers free ETH withdrawal via Mantle on the routes where the network is open, which is one of the cleaner cross-chain options when Mantle is up. Regulatory caveat: US residents have been blocked from Bybit since 2023, so a US reader will see the entire Bybit endpoint as effectively disabled even when the API flag is on. Plan accordingly via the Bybit withdrawal fees guide or open an account through https://yieldo.me/go/bybit?ctx=web_blog if your jurisdiction allows.

OKX — closed networks profile and honest framing

OKX has the single most-discussed outlier in the dataset: 34 OKTC routes that are permanently disabled. OKTC, formerly OKExChain, is a legacy chain OKX has sunset gradually since 2023. The routes still appear in the public currencies endpoint with fee = 0 and withdraw_enabled = false. Any third-party fee aggregator that doesn't filter on withdraw_enabled = true will display OKX as the most "zero-fee" exchange in the industry. That display is technically true and practically false: those routes don't move money. The OKTC block alone pushes OKX's aggregate disabled share into the 30–40% range, even though the rest of OKX's active-route portfolio is roughly average.

This is the right framing for OKX more broadly. The exchange is, by our measure, the most transparent about historical network deprecation — it leaves disabled routes visible in the API rather than silently removing them. That transparency penalizes OKX in any naive count of "total disabled routes," but it's the correct behavior from a downstream integrator's perspective. On the active routes, OKX offers some of the lowest practical fees in the dataset: USDT on Plasma at OKX clears at about $0.000059, effectively free. Communication via @okx is steady; the OKTC sunset has been announced multiple times.

If you need an exchange with deep current-generation network coverage and don't care about the historical OKTC tail, OKX remains one of the strongest fits. See the OKX withdrawal fees overview or open an account via https://yieldo.me/go/okx?ctx=web_blog when your target route is verified open.

Bitget — closed networks profile and honest framing

Bitget runs the most predictable closure profile in our snapshot. Maintenance windows are typically announced 24+ hours in advance, the share of structural permanent-off routes is small, and the exchange rarely closes routes reactively. The one persistent exception is BITLAYER for BTC: the zero-fee route is listed but disabled in our data and has been for the entire tracking window.

Bitget also holds a Protection Fund of $300M+, which functions as a hack-recovery reserve rather than a route-stability mechanism — different problem, separate context. Regulatory: Bitget never served US users and has restricted access in the UK and Russian Federation following EU sanctions packages. Outside those constraints, Bitget is the "quietly reliable" option in the dataset, with Polygon, Arbitrum, and BSC routes that stay open through most network events. See the Bitget withdrawal fees guide or open an account via https://yieldo.me/go/bitget?ctx=web_blog when your route is verified open.

Gate.io — closed networks profile and honest framing

Gate.io has the deepest catalog of any exchange in the dataset by an order of magnitude — well over 1,000 staking products and several thousand withdrawal routes. That breadth is double-edged. The per-route closure rate is roughly in line with the cross-exchange average (~30–40% withdraw disabled, ~60–70% deposit disabled in our snapshot), but the absolute disabled count is much higher simply because there are more routes to count. Reading the headline number as "Gate is the most restrictive" is wrong — the right framing is "Gate has the deepest long tail, and long-tail routes have higher base-rate closure probability everywhere."

A separate caveat that affects every Gate.io reader: the exchange lists a class of leveraged tokens (3L/3S/5L/5S and variants) that are technically derivatives, not self-custodial tokens. Their withdraw_enabled = false flag is structural — you cannot custody these instruments outside Gate.io because they don't exist outside Gate's product layer. Treat them as off-route rather than as freeze events. On the regular-token routes, Gate.io is the right pick when you need an unusual coin that nobody else carries. See the Gate.io withdrawal fees guide or sign up via https://yieldo.me/go/gate?ctx=web_blog when your route is open.

KuCoin — closed networks profile and honest framing

KuCoin is the textbook case of an exchange that survived a major breach and stayed honest about it. In September 2020 the exchange suffered a $281M hack, paused withdrawals across affected networks for an extended window, progressively re-enabled routes as funds were recovered, and ultimately reimbursed users in full by February 2021. That "wide freeze → slow re-enable → full reimbursement" sequence is one of two cases in the industry where users were made whole after a major breach (the other being BingX's September 2024 incident).

Today, KuCoin's per-route disabled share runs in the 25–35% withdraw range — roughly in line with the dataset median. The visible regulatory caveat: after a CFTC settlement in March 2024, KuCoin closed access to US residents. US readers will see effectively 100% disabled routes regardless of the API flag — that's jurisdiction policy, not technical closure, and it does not factor into our 28.9% / 56.2% headline numbers, which measure API-level state. See the KuCoin withdrawal fees guide or open an account via https://yieldo.me/go/kucoin?ctx=web_blog where allowed.

Binance — context-only

Binance is the largest exchange in the world by spot volume and runs a withdraw_enabled / deposit_enabled flag system substantially similar to the six exchanges above. Binance is not in our promoted route set, which is a structural decision rather than a quality judgement — we maintain a Binance provider for reference but do not include it in the live tracker's headline numbers. Where it's worth flagging Binance in this study: the exchange runs the industry-best closure-announcement cadence, typically posting 24–48 hours of advance notice for scheduled maintenance, and the Binance Academy documentation for wrong-network recovery scenarios is the most thorough public-facing material on the topic. We reference Binance in this study for cross-industry context but do not link out to its referral program.

Network-Level Analysis — Legacy vs New Networks

Reading the per-exchange profiles is one cut of the data. The other cut is by network, and it tells a slightly different story. Closures are not uniformly distributed across all networks — they cluster, and the clusters tell you which networks are robust and which are structurally fragile.

Why TRC20 and ERC20 stay open more often

TRC20 (Tron) is the single most reliably open network for USDT in our snapshot. Roughly 90%+ of TRC20 USDT withdraw routes are typically on across the seven supported exchanges. The reason is structural: Tron's on-chain cost is low, energy is rentable on a decentralized market, and the network has very few upgrade events that require precautionary freezes. Tron Proposal #104, dated 29 August 2025, dropped the energy price from 210 sun to 100 sun, halving the effective on-chain cost of a TRC20 USDT transfer. That structural update made TRC20 even more attractive for exchanges to keep open — the cost of running a TRC20 endpoint is now roughly half what it was. The ERC20 vs TRC20 vs BEP20 deep-dive covers the per-network economics in more detail.

ERC20 (Ethereum mainnet) is the second most reliably open network for stablecoins, with the major caveat that Ethereum mainnet upgrade events trigger broad precautionary freezes. The ETH Shanghai upgrade in April 2023 (activation of stake withdrawals) prompted most CEXs to pause ETH ERC20 withdraw and deposit endpoints for the first 24–48 hours post-activation. That pattern repeats with every Ethereum hard fork. Outside upgrade windows, ERC20 stays open.

BEP20 (BSC) sits in the same category. Stable in normal conditions, briefly fragile during major BSC upgrades like the mid-July 2023 Fast Finality migration, which prompted precautionary deposit pauses across several exchanges.

Why newer L2s see more closures

Newer L2 networks (Plasma, X Layer, Mantle, Linea, Scroll, zkSync Era variants, Base) see more closures for several structural reasons. They go through more upgrade events — each major protocol release is a candidate freeze trigger. They have more concentrated indexer infrastructure — an RPC issue at a single major provider can take all CEX deposit endpoints offline on that L2 simultaneously. And they have less inventory at the exchange — small hot-wallet balances on a niche L2 are a leverage point for arbitrage drain, so exchanges err toward closing rather than holding open.

Specific cases worth flagging: zkSync Era and Starknet variants see asynchronous batch processing, which makes live deposit indexing harder than on a normal EVM L2. Linea and Scroll go through more frequent client-software upgrades than older L2s. Base is structurally robust but young — less history to draw closure-frequency conclusions from.

USDT, USDC, and stablecoin route fragility

USDT is the most cross-chain stablecoin in the industry, with active routes on 15+ networks. Each exchange supports a different subset:

• MEXC supports USDT on 10+ networks
• Bybit supports USDT on 9–10 networks
• OKX supports USDT on 8–9 active networks (plus the 34 historically disabled OKTC routes)
• Bitget supports USDT on 7–8 networks
• Gate.io supports USDT on 12+ networks
• KuCoin supports USDT on 7–8 networks

This per-exchange variance is the structural reason the deposit/withdraw asymmetry is visible in the data. One exchange supports USDT deposits on Plasma but not withdrawals; another exchange does the opposite. Naive cross-exchange route planning that ignores network-level differences runs into a dead end quickly. The cheapest way to send USDT walks through the practical cross-exchange routing problem in detail. USDC behaves similarly, though its network catalog is narrower (USDC is concentrated on Ethereum, Solana, Polygon, Arbitrum, Base; less commonly on TRC20).

Why Deposits Get Disabled More Often Than Withdrawals

Coming back to the asymmetry. The most surprising single finding in this study is that deposits are disabled roughly twice as often as withdrawals. Aggregate snapshot: 56.2% deposit disabled vs 28.9% withdraw disabled on 25 June 2026. That is not a coincidence and it is not a transient artifact. It is structural, and it shows up consistently across days and across exchanges.

Liquidity rebalancing on the deposit side

The most common reason a deposit endpoint is offline is that the exchange has paused incoming flow on that route to rebalance hot-wallet inventory. Hot wallets are the working balances exchanges use to credit user deposits and fund user withdrawals. When the hot-wallet balance on a particular (coin, network) tuple gets too thin, the exchange has two choices: top it up from cold storage (slow, requires sign-off) or stop accepting new inbound flow (fast, low-cost). The default response is the second. Pausing deposits doesn't lose anyone money — it just delays the next inbound transfer. Pausing withdrawals does lose money, in the sense that it strands user funds and triggers support tickets. So the incentive structure is asymmetric: close deposits readily, close withdrawals only when forced.

Hot wallet vs cold wallet asymmetry

The deposit pipeline is a four-stage system: chain indexer (watch for incoming transactions to user addresses) → memo / destination-tag parser → hot-wallet crediting (move funds from the chain into the exchange's internal balance) → user-account balance update. Any of those four stages can fail independently. If the indexer falls behind, deposits pause. If memo parsing breaks (XRP, EOS, BNB Beacon Chain, TON), deposits pause. If hot-wallet crediting hits a signing-key issue, deposits pause.

The withdrawal pipeline is a two-stage system: sign outgoing transaction (from a known internal balance) → broadcast to the network. Fewer stages, fewer points of failure, more controlled flow. The exchange knows what it's sending and verifies it once before signing. The asymmetric stage count is the structural reason deposits fail roughly twice as often.

Exchanges err toward freezing inbound first

There's a secondary structural reason that compounds the infrastructure point. A single problematic RPC node on, say, BSC takes every BEP20 deposit endpoint offline simultaneously — USDT, BUSD, BNB, ETH on BEP20, and so on. One infrastructure issue cascades across 50+ coin-network combinations on the deposit side. The same RPC issue typically does not affect withdrawals on those routes, because the exchange is signing outbound from its own internal state rather than reading inbound from external indexer state. So a single deposit-side outage can spike the disabled count visibly, while withdrawals stay roughly stable. That cascading-failure pattern is one of the largest contributors to the persistent 2:1 deposit-to-withdraw asymmetry.

Five Causes of Network Closures You'll See in the Data

Reading the closure data across weeks, five recurring patterns emerge. None of them are exotic and none of them are evidence of villainy on the exchange side. They are the structural reality of operating a hot-wallet-backed CEX in a fragmented multi-chain environment.

Network upgrades and hard forks

Exchanges pre-emptively pause withdraw and deposit endpoints around blockchain upgrades to prevent replay attacks, double-spend, or stuck deposits during the activation window. Examples we've seen repeatedly:

• ETH Shanghai — April 2023. Activation of staked-ETH withdrawals. Most major CEXs paused ETH ERC20 routes for 24–48 hours post-activation.
• BSC Fast Finality — mid-July 2023. Migration to single-slot finality on BNB Smart Chain. Several CEXs temporarily closed BEP20 deposits.
• Tron Proposal #104 — 29 August 2025. Energy price adjustment from 210 sun to 100 sun. TRC20 routes stayed open, but the fee structure shifted underneath.
• Bitcoin Taproot activation — November 2021. Bitcoin mainnet routes saw 24–48 hours of precautionary pauses.

Roughly 97% of upgrade-related closures reopen within 72 hours. The pattern is predictable and announced in advance.

Exploit response and emergency hold

Exchanges reactively freeze routes after a security event — either their own incident or one on a counterparty (a bridge, a DEX, a custodian). Documented examples:

• Bybit Lazarus hack — 21 February 2025. ~$1.5B in ETH drained from a cold wallet. Wide freeze on ETH-network withdraw routes in the first 24 hours, full reopening in 72 hours, insurance-fund-backed user reimbursement.
• KuCoin hack — September 2020. $281M lost. Widespread freeze across affected networks, slow re-enable, full reimbursement by February 2021.
• Curve Vyper exploit — 30 July 2023. ~$70M lost across Curve pools. CEXs holding CRV on Curve-linked DEX routes temporarily paused CRV withdrawals.
• BingX hack — September 2024. $43–52M lost in a Lazarus-linked incident. Withdrawals restored within ~24 hours, deposits within 48, with full user reimbursement.

Exploit-response closures are typically sharp and fast — 24 to 72 hours. Exchanges are strongly incentivized to restore trust quickly rather than hold routes closed indefinitely.

Liquidity and inventory rebalancing

The single most frequent — and also the most invisible — closure cause. Exchanges auto-pause low-volume coin-network combinations when inventory dips below a threshold, when an arbitrage drain pattern is detected, or when a token has just printed an outsized run-up on a fast-listing exchange. These closures usually don't make announcement headlines because they are routine inventory operations. They are also the largest contributor to the headline 56% deposit-disabled share — exchanges pause inbound inventory aggressively to avoid having to top up hot wallets for thin routes.

Regulatory pressure and sanctions

Regulatory action triggers route closures for specific coins, specific pairs, or specific jurisdictions. The recent reference set:

• SEC vs Ripple settled — August 2025, $125M penalty. XRP routes had been periodically paused on multiple CEXs before settlement; XRP ETF approvals in November 2025 standardized XRP as "always open" afterward — a rare case where regulatory clarity reduced closure frequency.
• MEXC MX Holder suspension — 9 February 2026. KYC tightening prompted discontinuation of certain discount programs.
• MiCA enforcement in the EU (phased 2024–2026). Exchanges without a MiCA license freeze certain EU-region routes.

Regulatory closures are the most permanent of the five categories. They typically don't reopen.

Routine maintenance windows

The most boring category. Exchanges schedule maintenance on internal infrastructure — database migrations, hot-wallet rotations, signing-key updates — and pause affected routes during the window. These are typically announced 24+ hours in advance and last under four hours. Bitget and Bybit are the cleanest cases in the dataset for this pattern. If you see a route disabled overnight and your exchange's announcement page shows a maintenance window, that's almost always the explanation.

How to Check Network Status Before Transferring (4-Step HowTo)

The data above is useful in aggregate. For an individual transfer, the practical question is more focused: is the network I want to use open at the source and destination, right now? Four steps.

Step 1 — Check the live status page

Open /fees/network-status on Yieldo. The page displays current withdraw_enabled and deposit_enabled status across all supported exchanges, refreshed every 30 minutes. Filter by the coin you're moving (USDT, ETH, BTC, SOL, XRP, USDC are the most common). The page also surfaces recent transitions — if a route flipped off in the last hour, that's worth knowing before you start the transfer.

Step 2 — Verify withdraw_enabled AND deposit_enabled on both sides

A transfer requires withdraw open at the source AND deposit open at the destination, on the same network. Both flags need to be on. If you're sending USDT TRC20 from MEXC to Bybit, you need MEXC's USDT TRC20 withdraw flag = on AND Bybit's USDT TRC20 deposit flag = on. Either being off blocks the transfer. The fees hub page shows aggregated per-coin views; per-exchange views are at the exchange-specific fee pages like the Bybit fees overview and the OKX fees overview.

Step 3 — Pick a fallback network if your first choice is closed

If your preferred network is closed at either end, pick a fallback. For USDT the typical fallback ladder is TRC20 → Polygon → BEP20 → Arbitrum → Solana, ordered roughly by cost. For ETH it's Arbitrum → Optimism → Base → Mantle → ERC20 (most expensive). For BTC it's Lightning → Liquid → BTC mainnet. The how to transfer crypto between exchanges guide walks through the practical fallback decision tree.

Step 4 — Send a small test transfer before the full amount

If the route shows open and you've verified both flags, send a small test transfer first — the minimum withdrawal amount on most exchanges is roughly $10–$25 worth. Wait for the destination to credit. If the test arrives in your destination balance, the route is genuinely operational and you can send the full amount. If the test stalls beyond the network's typical confirmation window (10 minutes for most L2s, an hour for Bitcoin mainnet), pause and check the route status again. The avoid wrong-network mistakes piece covers what to do if a test does land in a route that later turns out to be down.

What to do if you're mid-transfer right now

If you've already pressed send and the destination shows deposits disabled, the action steps are time-sensitive. The sister piece Withdrawal Network Disabled: What to Do covers the recovery workflow — in most cases funds are recoverable, but the procedure depends on the exchange, the network, and how quickly you flag the case to support.

Final Verdict — Closed Routes Are a Feature, Not a Bug

On any given day, ~28–30% of crypto withdrawal routes and ~55–60% of deposit routes are disabled across major CEXs. That is the structural backdrop of cross-exchange transfers in 2026, and it is not going away. It is the predictable consequence of an industry that operates hot-wallet-backed exchanges on top of a fragmented multi-chain stack, with asymmetric infrastructure on deposits vs withdrawals and asymmetric incentives on which side to pause first. The headline numbers will move — an Ethereum upgrade can push withdraw-disabled into the 35% range for 48 hours, a quiet inventory week can pull deposit-disabled below 50% — but the structural shape is stable.

For day-to-day users

Bookmark /fees/network-status and check it before any cross-exchange transfer. Don't trust the network dropdown on either exchange's withdraw modal in isolation — the source side might show TRC20 as available while the destination has it disabled, and you only see that on a tool that compares both sides. For routine USDT moves, TRC20 is almost always your safest bet; for ETH, an L2 like Arbitrum or Base if both sides are open; for BTC, Lightning if available. Honest CTAs for day-to-day routing: https://yieldo.me/go/mexc?ctx=web_blog for the broadest catalog, https://yieldo.me/go/bybit?ctx=web_blog for the most predictable closure profile, https://yieldo.me/go/okx?ctx=web_blog for the most transparent treatment of historical routes.

For arbitrage and route-sensitive traders

Closed routes are the constraint that makes cross-exchange arbitrage work. If every route were always open, arbitrage spreads would collapse to fee + slippage in milliseconds. The fact that one exchange has a route closed while another has it open is what creates the persistent price differentials that arbitrage traders capture. Bake the /fees/network-status JSON feed into your route-availability checks before placing the trade — a 3% spread on a closed route is a 0% spread in practice. https://yieldo.me/go/bitget?ctx=web_blog and https://yieldo.me/go/gate?ctx=web_blog cover the long-tail altcoins where route arbitrage is most active.

For long-term holders

If you hold funds on a CEX, periodically verify that withdrawals are open on your preferred network. The wider lesson from the dataset is that withdrawal availability is not guaranteed continuously — an unexpected exploit response, a sudden regulatory action, or a maintenance window can lock you out for hours or days. Cold-storage exposure is the structural answer for long-term holdings; for funds that need to stay liquid on exchange, diversify across at least two of the supported set: https://yieldo.me/go/mexc?ctx=web_blog, https://yieldo.me/go/bybit?ctx=web_blog, https://yieldo.me/go/okx?ctx=web_blog, https://yieldo.me/go/bitget?ctx=web_blog, https://yieldo.me/go/gate?ctx=web_blog, https://yieldo.me/go/kucoin?ctx=web_blog.

For parallel research on the same kind of structural-pattern-in-data approach applied to staking, see the cross-link do exchanges cut APY after listing. For methodology and data provenance documentation, see /about/data.

Frequently Asked Questions

How many crypto withdrawal routes are disabled right now?

Across the seven supported exchanges in our snapshot, roughly 28–30% of withdraw routes and 55–60% of deposit routes are disabled at any random sampling moment. On the 25 June 2026 reference snapshot the exact figures were 28.9% and 56.2%. Live current values shift on every 30-minute refresh — check /fees/network-status for now-state. The ranges are stable across recent weeks; the point estimates move.

Why do exchanges disable deposits more often than withdrawals?

The deposit pipeline has more failure modes than the withdrawal pipeline. Deposits need a working chain indexer, memo parser, hot-wallet crediting service, and internal balance update — four stages, any of which can fail. Withdrawals need a signing service and a broadcaster — two stages. Add that exchanges are biased toward closing inbound flow first because doing so doesn't strand user funds (closed withdrawals do). The combination explains the persistent ~2:1 deposit-to-withdraw closure ratio.

Which exchange has the most closed networks?

By per-route share, OKX and Gate.io typically run the highest disabled rates in our snapshot. OKX's share is heavily influenced by the 34 OKTC legacy routes that have been permanently disabled since the OKTC sunset. Gate.io's share reflects its deep long-tail catalog — more routes means more rare-altcoin routes that have low base-rate operating probability. Per-route closure rates are roughly in line with the dataset average for both exchanges. By absolute count of disabled routes, the ordering shifts with catalog depth. See the per-exchange profiles above.

Are network closures permanent or temporary?

Most closures are temporary. Upgrade-related closures typically reopen within 72 hours. Exploit-response closures reopen in 24–72 hours as exchanges restore trust. Liquidity closures often lift within hours when inventory rebalances. The permanent closures in the dataset are the structural sunsets — OKX OKTC's 34 routes, MEXC ZKSYNC LITE for ETH, Bitget BITLAYER for BTC — and regulatory closures, which rarely reopen. The remaining 90%+ of closures you'll see in the live tracker are temporary.

How can I check if a network is open before withdrawing?

Open /fees/network-status and filter by your coin. Verify that BOTH withdraw_enabled at the source exchange AND deposit_enabled at the destination exchange are on. Send a small test transfer before the full amount. If the test arrives within the network's typical confirmation window, send the rest. The full step-by-step is in the HowTo section above; the how to transfer crypto between exchanges guide covers the practical decision tree.

What should I do if my withdrawal is disabled right now?

If you can't initiate the withdraw, switch to a different network on the source exchange and verify the destination supports it. If your transfer was already broadcast and the destination has deposits disabled, the action workflow is in the sister piece Withdrawal Network Disabled: What to Do. In most cases funds are recoverable, but the recovery procedure depends on the specific exchange and network involved.

Does Yieldo's tracker work in real-time?

The tracker refreshes every 30 minutes via official exchange APIs. That cadence is the closest to real-time that's practical without hammering exchange endpoints and risking rate-limit issues. Within any 30-minute window, the displayed state can lag actual exchange state, so for time-sensitive transfers we recommend double-checking the source exchange's own withdraw modal before sending. For aggregate research and pattern recognition, the 30-minute cadence is more than sufficient.

Why does Bybit / MEXC / OKX close so many networks?

Bybit closes networks less often than the dataset average, mostly via scheduled maintenance windows announced in advance. MEXC closes more frequently because its deep low-cap altcoin catalog has many narrow routes that auto-pause on liquidity thresholds. OKX's headline number is inflated by the 34 permanently disabled OKTC legacy routes; its active-route closure rate is roughly average. None of these patterns indicate negligence — they reflect different catalog compositions and different risk-management postures. The per-exchange profiles above cover the structural detail.

Is this data publicly available and reproducible?

Yes. The full dataset is available via the /fees/network-status live page, the /fees/network-status.json machine-readable feed, and the /fees/network-status.csv full event-history export. All under CC BY-NC 4.0 — please cite Yieldo as the source. The methodology and provider list are documented at About the data. Independent researchers, journalists, and developers are welcome to use this dataset; for v2 longitudinal queries, the event log has been accumulating since 25 June 2026 and grows by tens to hundreds of events per day across the seven supported exchanges.

Is the network status tracker available in the United States?

The tracker itself is publicly accessible from any jurisdiction. Access to the underlying exchanges varies. Bybit, MEXC, OKX, and Gate.io restrict US-resident access in most cases; KuCoin closed US access after the March 2024 CFTC settlement; Bitget never served US users. US readers can still use the tracker for educational and research purposes and for transferring funds among any exchanges they do have access to, but should consult their own legal counsel before opening accounts on any non-US-licensed exchange. None of this article constitutes financial or legal advice.

Limitations, Author, and Disclaimer

Data limitations (v1 snapshot)

This study is a snapshot, not a longitudinal time-series. The 28.9% / 56.2% headline figures are from the 25 June 2026 reference snapshot; live values shift on every 30-minute refresh. The event log (network_availability_events) began accumulating on 25 June 2026 and needs several more weeks of data before per-route freeze-frequency and mean-time-to-recovery findings can be published responsibly. A stale-data guard within our aggregator may flip a small fraction of routes to withdraw_enabled = false on ingestion-gap edge cases; those flips are not written to the event log but do appear in snapshot numbers. Use the ranges (~28–30%, ~55–60%) rather than point estimates when interpreting trend.

Risk warning

Cross-exchange transfers carry irrecoverable failure modes when networks are closed or addresses are mis-formatted. Always verify both withdraw_enabled at the source and deposit_enabled at the destination before sending. Test transfers are mandatory for any move above a few hundred dollars. Exchange custody is inherently riskier than self-custody — funds on a CEX can be locked by regulatory action, security incidents, or operational events even when withdraw flags appear on. This article is not financial, tax, or legal advice; consult a licensed professional before making transfer decisions involving significant capital.

Author

Written by Eugen Voyager — crypto analyst and founder of Telochain blockchain, GameFi project @telomeme, and the @tonsdot Telegram channel covering exchange operations, DeFi, and market dynamics. Practical experience operating blockchain infrastructure and tracking exchange behavior underpins the methodology used in this study.

Disclaimer

This article contains affiliate links. Yieldo may earn a commission at no extra cost to you when you sign up for an exchange via one of the referral links above. Affiliate relationships do not influence the methodology, data, or per-exchange framing in this study — the dataset is the same whether you click a referral link or not, and per-exchange numbers come from the public APIs of each exchange, not from the exchanges' sales teams.

Last updated: 30 June 2026.