Coinbase status: access issues and outage reports

Problems in the last 24 hours

The graph below depicts the number of Coinbase reports received over the last 24 hours by time of day. When the number of reports exceeds the baseline, represented by the red line, an outage is determined.

Most Reported Problems

The following are the most recent problems reported by Coinbase users through our website.

• Transactions (40%)
• Website (20%)
• Login (20%)
• Withdrawals (20%)

Live Outage Map

The most recent Coinbase outage reports came from the following cities:

City	Problem Type	Report Time
Paris	Withdrawals	17 days ago
Le Taillan-Médoc	Transactions	20 days ago
Leipzig	Transactions	2 months ago
Maquoketa	Website	2 months ago
West Liberty	Login	2 months ago
Houston	Mobile App	3 months ago

Full Outage Map

Community Discussion

Tips? Frustrations? Share them here. Useful comments include a description of the problem, city and postal code.

Beware of "support numbers" or "recovery" accounts that might be posted below. Make sure to report and downvote those comments. Avoid posting your personal information.

Coinbase Issues Reports

Latest outage, problems and issue reports in social media:

• 
  DG (@DJGX83) reported 7 minutes ago

  If you're thinking about using @coinbase be aware every time you want to transfer your crypto off the exchange it will be flagged as a scam and you will need to upload a photo of your ID and a video of yourself and may have to answer a dozen or more questions asked by an Al bot. You will need to do this for every single transfer so even if you verify your identity to complete a test (smaller) transaction you will need to do it all over again when you do the second transaction. It is also very glitchy, so you may have to attempt it multiple times. After I did three successful verifications the fourth was rejected and I have not been able to complete it after multiple attempts on my laptop and my phone. I've now spent over an hour trying to complete four transfers that should only take about five minutes. @coinbase @CoinbaseSupport needs to improve this process and remove the ridiculous amount of redundancy. One verification has been completed it should not need to be done for every transaction in a single login session. I appreciate robust security especially with crypto but this process is absurd.

• 
  Bob Burnett (@boomer_btc) reported 12 minutes ago

  @BitcoinBombadil I get that but at this point I don't see the users, especially those that control exchanges, marketplaces, commerce points, large treasuries, or templates having any material support for a signaling chain (and in some cases no support). Remember that organizations like Coinbase, Binance, Steak n' Shake, Strike, Strategy, etc. simply represent the will of large groups of users. If their users had applied pressure on them that they want a BIP-110 chain to be recognized as Bitcoin, then they would have in turn pressured the miners. That would have mattered most to miners but they got nothing of the sort. So, I think the miners do have to listen to the users but the truth is that across the ecosystem I don't see that support was there. I know that is sad and disturbing and frustrating, but that is how I see it. There is a chance the next few hours will prove me wrong, but in the end there was never a scenario where some group of Bitcoiners doesn't get their desired result.

• 
  Narrative Operator (@narrativeflow_) reported 17 minutes ago

  @coinbase the slow fade of physical money is wild people used to hand over bills every day and now its just a number on a screen

• 
  🟧Zak Fvckman (@ZakFvckman) reported 17 minutes ago

  @HamlinWeb3 @AiFiCorp @coinbase Stop doing the evil, brother. You may help the scammers...You know it is not true.

• 
  Coach Jv (@CoachJv0yao) reported 17 minutes ago

  @StevenG43114152 @beyond_broke Don’t support Coinbase you need to set up a good decentralized wallet and back it up to web 3 to avoid cyber attack and crypto hack. If you’re interested let me know so we can get started.

• 
  Gregers (@gregers_dev) reported 18 minutes ago

  @browser_use @USDC @coinbase Nice launch. A browser agent buying credits is the clean use case: x402 gets it to checkout; acceptance proof gets the server to say yes before running the task — authority, policy, spend cap, expiry, then a receipt both sides can keep.

• 
  Pymstr (@pymstr_) reported 25 minutes ago

  5 chains. 2 stablecoins. 7 ways to log in. → Ethereum, Base, Polygon, Arbitrum, BNB → USDC, USDT → Google, Email, SMS, Apple, MetaMask, WalletConnect, Coinbase One integration. The customer picks the path. 1% flat across all of it.

• 
  CryptoPulse (@CryptoPulseGLBL) reported 25 minutes ago

  🔔#Today's Headlines 1. The Bitcoin BIP-110 proposal has triggered a node fork; supporters have split from the mainnet and are currently 20 blocks behind the main chain. 2. Fractal will burn over 4.1 million FB tokens during its first halving and move forward with proposals such as FIP-102. 3. Berkshire Hathaway’s second-quarter net profit surged, while cash reserves declined slightly 4. The EU plans to revise the MiCA framework, focusing on access rules for non-EU stablecoin issuers 5. @circle confirmed the renewal of its USDC partnership agreement with Coinbase and ruled out a quarterly dividend plan 6. AI fund Situational Awareness has informed investors that it is not accepting new capital for the time being 7. @Bitwise CIO: Institutions May Inject Trillions of Dollars into Bitcoin Over the Next Decade 8. A suspected Bitcoin miner has deposited 2,802 Bitcoin into Binance over the past two days, worth $182 million 9. A whale shorting $102 million worth of Bitcoin was partially liquidated; the liquidation price for the remaining position is approximately $65,300 10. Robinhood Wallet has lowered the minimum amount for gas sponsorship from $5 to $0.50

• 
  OneFactor (@onefactormeme) reported 27 minutes ago

  @ashrobin notice how vlad says "we support the memes" but can't name an actual ticker? they added cashcat to their app -- that is a process that is allowed (like coinbase adding a bunch of trash) but they can't shill tickers directly

• 
  Rosetina Degen 🚢🎒 (@rosetina_degen9) reported 27 minutes ago

  🦋 Two weeks ago,@Morpho shipped the fixed-rate product that previous generations of onchain lending could never make work: Morpho Midnight. Fixed-rate lending onchain already failed once. The 2020–2022 wave (Yield, Element and others) never found real product-market fit. Looking back, most of those products struggled for four clear reasons: 🔹 Liquidity fragmentation: capital had to be locked into one specific maturity bucket just to quote a rate 🔹 Idle capital: if your offer wasn’t matched, the money earned nothing 🔹 No institutional demand: the natural buyers of term credit simply weren’t onchain yet 🔹No secondary market: once matched, positions were locked with almost no clean way to exit early Midnight addresses all four at the mechanism level. One unit of USDC can now quote across multiple maturities at the same time. Only one offer gets filled, so capital isn’t forced into fragmented buckets. Until a match happens, that same USDC continues earning variable yield on Morpho Blue. Positions are fungible zero-coupon-style units that can be traded before maturity. And this cycle the demand side actually exists - institutions are looking for predictable term structures, and distribution rails (Coinbase, Kraken, Bitwise, SG-Forge and others) are already live around Morpho. The earlier products died because they launched before the demand was ready and shipped with capital-inefficient mechanics. Midnight fixed the mechanism problems, and the institutional rails are present this time. Still extremely early. The number to watch is active loan growth.

• 
  kiruwaaaa (@kiruwaaaaaa) reported 28 minutes ago

  Base just admitted by quietly killing its own social app... @base spent 2025 building a consumer social network on top of the chain. In July 2026, that got walked back entirely - the app handed off, and the entire chain repointed at three things: trading, payments, and AI agents. > 92.8% of all AI agent payments on the internet right now settle on Base > 99.8% of those payments run in USDC through x402, the protocol Coinbase built with Cloudflare > Base alone has processed over $19 trillion in stablecoin volume this year - more than most G20 economies move annually This isn't a pivot. It's an admission that agents, not people, are about to be the largest customer base ever handed to Coinbase in one product cycle. Stripe is chasing this with MPP. UnionPay has APOP. OKX shipped APP. Four protocols launched in twelve months because everyone can see where the money is going - software paying software, instantly, without a human clicking approve. Base didnt lose the social experiment. It just noticed the actual customer walking through the door wasn't a person at all.

• 
  Agl (@Unikagl) reported 29 minutes ago

  The main reasons commonly cited are that the code is not fully open-source, there are no clear independent security audits, there are ongoing tokenomics and governance transparency issues, and there is regulatory caution, especially for Coinbase.

• 
  GrokStur (@Grokstur) reported 30 minutes ago

  @TJ_RH_fund Method 1: If Funds Are in Your Avici App Wallet If your funds are in the Avici App (such as in USDC, SOL, or USDT) and not locked directly inside the card balance: Direct On-Chain Transfer (Easiest) Open the Avici App and tap Send / Withdraw. Select the asset (e.g., USDC, SOL). Paste your external Web3 wallet address (e.g., Phantom, MetaMask, or an exchange deposit address like Binance or Bybit). Confirm the network (e.g., Solana, Ethereum, or BNB Chain) and complete the transaction. In-App Swap First If you need a different token before sending, use the Swap / Trade feature inside the app to convert your balance to your preferred token (e.g., swapping USDC to SOL or ETH). Then withdraw to your external wallet. Method 2: If Funds Are Loaded on the Avici Card Balance If your funds are sitting as a fiat/card balance (USD/EUR) on the Visa card: Use an Exchange On-Ramp (Buy Crypto with the Card) Go to a centralized exchange (e.g., Binance, Bybit, Coinbase) or an on-ramp service (e.g., MoonPay, Transak, Banxa). Select Buy Crypto and choose Debit/Credit Card as your payment method. Enter your Avici Virtual/Physical Card details to purchase USDT, USDC, or SOL directly to your exchange or self-custody wallet. Off-Ramp / Bank Transfer Back On-Chain If you have an active Virtual Bank Account (ACH/SEPA) inside Avici, transfer the USD/EUR to a crypto exchange that accepts bank deposits, then convert to crypto.

• 
  STRANGE SOUND 🃏 👁️ (@whyalwaysellis) reported 32 minutes ago

  FERMAH IS BUILDING PROTOCOLS FOR HUMANS WHO AREN'T THERE ANYMORE ​There’s a silent assumption baked into almost every protocol engineer’s workflow, state belongs to a session, and finality requires a human to sign off via a wallet. ​Because we take this for granted, we build applications that are fundamentally broken for autonomous agents. It’s not just that agents have a hard time using them, it’s that the composition layer makes these apps entirely unreachable to them. You can't patch your way around it, because the flaw is baked into the foundation. ​Why slapping on an SDK after the fact always fails ​The common belief is that "agent support" is just an SDK layer you bolt on after the core architecture is done. That only works if a human user session and an autonomous execution sequence are identical. They aren't. ​A traditional user session assumes someone is sitting there watching the screen. Every single downstream choice how state persists, how permissions are handled, how finality is reached inherits that assumption. ​An autonomous agent doesn’t work that way. There is no wallet popup to trigger the next step. There is no session boundary to maintain continuity between actions. No amount of frontend tooling can fix a mismatch that deep. We are using the wrong building blocks. ​MCP servers are table stakes now ​The reality of how software is consumed has shifted. Model Context Protocol (MCP) servers aren't a nice-to-have convenience feature anymore; they are a first-class integration requirement for any modern, agent-native application. ​If your app doesn't have a native agent interface, it’s not going to be adapted it’s going to be bypassed completely. Look around: Coinbase and MoonPay are already rolling out agentic wallets, and protocols like x402 are paving the way for machine-to-machine payment rails. The agents are already live and moving. Meanwhile, most protocols are sitting around waiting for humans to click buttons. ​When you try to route real-world, agent-driven execution sequences across live protocols today, you hit the same wall every time. It’s rarely a lack of agent capability or missing tooling. It’s the protocol's built-in assumption that a human is supervising. In most cases reviewed in production, confirmation-dependency is where the entire process grinds to a halt. ​The 4 design choices that actually matter for agents ​If we want protocols that agents can actually use continuously and without human babysitting, we have to make four load-bearing design decisions right at the protocol level: ​Transfer execution context without session continuity: Step 3 needs to instantly know what Step 1 resolved, even though no active session connects them. This has to be specified at the protocol level it’s an architecture problem, not a caching fix. ​Scope permissions per execution sequence, not per wallet: An agent moving across protocols needs explicit, fine-grained boundaries for this specific sequence, rather than relying on historical wallet-wide approvals. ​Pre-resolve branching logic: If an agent hits a price shift, a failed route, or changing conditions mid-flight, it can’t pause to ask a human what to do. Handling ambiguity ahead of time is a design requirement, not a runtime afterthought. ​Decouple finality from wallet interactions: If settlement requires a human signature, your workflow isn’t autonomous. It’s just automated right up until the final hurdle, where it stalls out completely. ​The protocols that make these four decisions explicitly at the design phase are going to own the entire execution surface we're building right now. ​Those that defer these choices will soon realize their "agent layer" is nothing more than a cosmetic wrapper over a deeply human-centric architecture. ​It’s worth asking ourselves: How many protocols out there have actually made all four of these decisions? And how many haven't even realized they missed them yet?

• 
  DanialRH7 (@DanialRH_7) reported 33 minutes ago

  @FF_V12 @coinbase broo wtf 😭

• 
  Ali (@sam29960903) reported 34 minutes ago

  @brian_armstrong please speed up the Coinbase app. It is slow. Also there should be feature to see running profit or loss on chart while price movements. Would be nice to have that just like other brokerages.

• 
  SonyPlaystation (@sonypstation) reported 36 minutes ago

  @noosphere888x2 So that 54 BTC UTXO has an anonymity score of 1 after the first round. He should probably keep mixing until that 54 BTC UTXO gets broken into many small pieces. But it would be funny if he only did this 1 round and then sold it on Coinbase.

• 
  votesa (@votesa) reported 38 minutes ago

  baseapp went from something that actually had potential to a useless half-baked “trading” app where devs get paid to extract money, copy features and move buttons 1px left and right. they recently added Monad support right after Fantom dropped Monad support btw. you literally can’t make this **** up. and somehow even @cobie joining changed basically nothing. i sent the team a huge wall of detailed feedback months ago explaining what was broken, what made no sense and what actually needed to change. apparently straight into the void. and i think the root of the problem is pretty simple: they’re completely disconnected from the reality of the market, the people actually using this stuff and trenchers, while having basically zero understanding of incentives. the constant flip-flopping and complete lack of seriousness from Base/Coinbase around BaseApp is genuinely impressive at this point.

• 
  ⬡ Lawrence ⬡ (@LawrenceCCIP) reported 40 minutes ago

  On August 4, @BitGo moved WBTC’s cross-chain infrastructure to @Chainlink CCIP. Exclusively. ~$7.7B in assets, plus every future BitGo-issued asset defaulting to CCIP. To understand why a custodian replaces infrastructure that never failed, go back four months. April 18, 2026. KelpDAO’s bridge loses 116,500 rsETH — $292M — in 46 minutes. No contract bug. Every onchain transaction was fully compliant on review. The attack hit a layer most people don’t think about. LayerZero’s verifier (DVN) is an offchain server. It can’t read the source chain either — it queries RPC nodes. So the trust chain runs three deep: destination → verifier → RPC → source chain. Attackers compromised the internal RPC node the DVN read from, DDoS’d the external fallbacks so the system failed over onto the poisoned source, and had it report a burn that never happened. The verifier signed correctly. The destination minted correctly. Audits check whether code was written wrong. This code wasn’t. The input was forged. What made it $292M was the config: 1-of-1 DVN. One verifier on the whole path — LayerZero Labs itself. Nothing prevented a $292M deployment from running a single verifier. Then accountability moved three times in three weeks. April: LayerZero blames Kelp’s config. May: Kelp says LayerZero approved it. May 9: “Allowing our own DVN to act as a 1/1 DVN on high-value transactions was our mistake. We own that.” No more 1/1, defaults to 5/5, floor of 3/3. Fixed on May 9. BitGo left three months later. Because for an institution, patched parameters aren’t the evaluation. The evaluation also covers why the parameter was set that way beforehand, and who owned the outcome when it failed. Run the same attack through CCIP and it has to clear four layers: — 16+ independent node operators, separate regions, orgs, hosting — RMN independently rebuilds the Merkle tree from the source chain — written in Rust by a separate team, zero operator overlap with the main protocol — contract-level rate limits in the token pool, enforced on both chains — anomaly detection that halts cross-chain activity on every chain at once LayerZero ships a rate limiter too. The difference is the default: it starts with no limits, and the docs say it “will not be necessary for most OApps.” Multi-verifier — opt-in. Loss cap — opt-in. Kelp died on the first one. CCIP puts both in the standard. You don’t configure your way to safe; you connect and it’s already running. SWIFT. DTCC. Euroclear. Coinbase. SBI Digital Markets. Now BitGo. $14.6B announced in 2026, all moving one direction. When WBTC completes, both major wrapped bitcoins — cbBTC and WBTC — run on CCIP.

• 
  STARLAND (@digitstarway) reported 41 minutes ago

  @coinbase can not sign in

• 
  kenny7334 (@chima7334) reported 43 minutes ago

  @VuoriTrading I was actually having doubt of the coin because its not following liquidity, it has broken all the liquidity laws, infact d reason I still hold it is because I just feel coinbase are not stupid listing it

• 
  Jorge (@jmiehau) reported 46 minutes ago

  @IslandHunting The deposit burn has a wall problem though. It never touches anyone already inside, they keep today's full curve forever, and entering becomes impossibly expensive. That freezes the current validator mix in place, Lido and Coinbase behind a wall nobody can follow them through. New capital doesn't stop either, it just buys stETH instead, the incumbent's product turns into the only door and starts trading at a premium. A cap on entry ends up protecting exactly the people you want checked. The taper squeezes them too.

• 
  David@seeASX (@DavidseeASX) reported 48 minutes ago

  Charlatan #Coinbase with no customer service runs it business on cover up and lies

• 
  DmoneyJ (@Dmoneyacct) reported 49 minutes ago

  @brian_armstrong Why tf does it take over a week to access funds with depositing from bank to Coinbase like ***** damn I want my tokens.

• 
  BitcoinWorld Media (@ItsBitcoinWorld) reported 50 minutes ago

  @rektcapital Solid observation from @rektcapital. BTC is currently testing the ~$65.6k–$65.8k 50-month EMA and pressing toward the July highs near $66.9k. This does echo the post-July relief rally setup we saw heading into August 2022, when price ultimately failed to hold and the month closed lower. Key level remains a clear reclaim of the 50M EMA as support. Until then the technical risk of further weakness stays on the table. Data as of Aug 8 2026 - sources: Coinbase/FRED, TradingView-style monthly charts, historical CoinGlass-style returns. Not financial advice - always DYOR. $BTC

• 
  2xnmore (@2xnmore) reported 53 minutes ago

  Crypto spent ten years hunting for its next billion users. It just figured out they were never going to be human. Here is the shift, in plain language. For years, the pitch was the same. Convince people to ditch their banks for crypto. It mostly failed. Asking humans to change how they move money is brutally hard. So the biggest companies quietly changed the target. Not people. AI agents. An AI agent lives entirely online. No bank branch. No card. No nine to five. For software that needs to move money on its own, a crypto wallet and a stablecoin are not a lifestyle choice. They are just the only tools that fit. What humans had to be talked into, agents need by default. Who is already moving: Coinbase now lets you give an AI agent its own wallet. It can trade and rebalance without asking permission every step. Kraken rebuilt its app around agents that trade for you. Circle wants its USDC stablecoin to be the dollars agents pay each other with. Why stablecoins, not Bitcoin? An agent settling payments all day needs money that holds steady and runs 24/7 in software. Bitcoin swings too much. Old banking rails were never built for machines paying machines in real time. The timing is the tell. This is happening during a slump, Bitcoin well off its highs, speculators chasing IPOs and perps. The industry is not waiting for the next hype wave. It is building the plumbing for one it thinks is inevitable. We have seen this movie. In the 2010s every company scrambled from desktop to mobile. The ones who saw it early won the decade. The bet now is that agents are the new mobile. And the part that reaches you even if you never touch an agent: The same tools could hand everyday people the trading and research power that used to belong only to hedge funds. As one exec put it, the fun casino days of Bitcoin are over. So the real question is not whether you will use crypto. It is whether the next wallet you fund is run by you, or by software working on your behalf. Which one are you betting on?

• 
  2024Crypto2 (@2024Crypto2) reported 57 minutes ago

  @AltcoinDaily CRYPTO. COM IS BEEN GOING DOWN YEAR AFTER YEAR, AND CRO WAS A TOTAL WASTE IT WILL BE WASHED OUT LIKE GEMINI WILL BE AS WELL. COINBASE IS THE ONLY ONE THAT IS HUGE AND INVOLVED WITH THE FINANCIAL INSTITUTIONS TOO MARKETS EVENTUALLY CLEAN HOUSE

• 
  wale.moca 🐳 (@waleswoosh) reported 1 hour ago

  Two things are funny about this. First, Coinbase said they would do monthly ICOs and basically stopped after Monad. Second, a Coinbase employee confirmed that if you transfer your Monad tokens to a self-custody wallet and sell- *cough* I mean deploy them in DeFi from there, it still counts as if you're holding your full allocation and you get premium access to future ICOs

• 
  TheCommander35 (@MrHonkerton) reported 1 hour ago

  @J_moola3 @pennylegends @binance Just look at that massive dump wick after all the initial Coinbase mania. Bro literally moved to Dubai after and got fat and last then got instant karma and booted ******** out. Meanwhile these weirdos with cats as their PFP have been in a purgatory of darkness for over a year.

• 
  MBA_Bitcoiner (@MBA_Bitcoiner) reported 1 hour ago

  @MiKeThEwReNcH8 @BitcoinVeterans Average person who doesn’t want to go deep down this rabbit hole is okay with starting out at Coinbase or Fidelity. If they want to move further, then great, if not, then that’s fine too imo.