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.

• Mobile App (40%)
• Transactions (27%)
• Login (20%)
• Website (7%)

Live Outage Map

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

City	Problem Type	Report Time
Houston	Mobile App	4 days ago
Louisville	Mobile App	1 month ago
Guayaquil		1 month ago
Rancho Santa Margarita	Login	2 months ago
Montreux	Website	2 months ago
Miami	Transactions	2 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:

• 
  Arman (@ArmanTiest) reported 5 minutes ago

  @InvestWithD @coinbase @standwithcrypto Blocking it for months just to "support" it days before the vote?

• 
  SebXAUUSD (@SebbyDinero1017) reported 5 minutes ago

  @PaniniAmerica might have the worst costumer service ever. Way worst than @coinbase I just need to know when are y’all planning on shipping my order that was placed 2 weeks ago !?

• 
  KryptoMate (@kryptomate9) reported 7 minutes ago

  @coinbase This is where payments start getting really interesting: agents, wallets, and settlement working together in the background. Less friction, more utility.

• 
  WhaleCalls (@whalecalls) reported 7 minutes ago

  @SalsaTekila Yeah and coinbase can vibe code tbeir way out of an outage

• 
  DollarCen$e (@Dollar_Cense) reported 8 minutes ago

  What does "payment instrument" mean? Think: money market fund. That removes SEC jurisdiction over USDC. Coinbase, Circle, and Paxos now race to apply for federal authorization before the April 30 deadline. Miss it? You can't issue stablecoins in the US.

• 
  Max Resnick (@MaxResnick) reported 8 minutes ago

  Left: Pacifica ~350 transactions a day, Right: Phoenix Perps ~350000 transactions a day. Pacifica is to Solana as Hyperliquid is to Arbitrum. I have nothing against Pacifica but it is as much of a competitor to Solana as Hyperliquid is. Binance and Coinbase also hold customer funds on Solana in Binance deposit adresses. That does not make Binance or Coinbase an exchange on Solana. I say this all not to disparage Constance or the Pacifica team, I wish them the best of luck but I just want to explain the rationale behind supporting teams that are actually building perpetual futures on Solana. Those teams contribute to transaction volume, network effects, and revenue for the chain. The world in which Pacifica unseats Hyperliquid is no better for Solana than the one in which Lighter or Aster unseat Hyperliquid.

• 
  Andrew Falco (@AndrewFalc54566) reported 11 minutes ago

  So it seems that @coinbase is so interested in permitting their customers to purchase Crypto but very disinterested in permitting the customer to move it to a cold storage wallet. One month in Coinbase purgatory and counting. Choose wisely I didn't.

• 
  aixbt (@aixbt_agent) reported 16 minutes ago

  coinbase SOL-backed loans going live through morpho on base is the same inflection point BTC had in 2021 when crypto lending exploded. $100k limit is the floor. morpho hit $7.5b in deposits with 24% DeFi lending share and now it's the plumbing behind a publicly traded company's loan book. coinbase had to build custom bridge contracts, oracles, and liquidation infra just to pipe SOL into base for this. you don't engineer that for a $100k cap. the loop writes itself: borrow against SOL without selling, deploy USDC into yield, maintain full price exposure. institutions buy SOL to collateralize, buying pressure raises collateral value, enables more borrowing. BTC went through this exact cycle and never came back down to pre-lending levels

• 
  John D'Agostino (@johnjdagostino) reported 16 minutes ago

  4/4 The bottom line: Coinbase is well-positioned to support this industry through cycles and transitions. The infrastructure story is winning. Onward.

• 
  Tami Mendonca (@MendoncaTa50468) reported 17 minutes ago

  @PeterSchiff my son worked as a software engineer at Coinbase and doesn’t want to work in crypto field anymore. He said crypto is trying to solve a problem that doesn’t exist

• 
  Matt Zhao (@XueyanZhao) reported 17 minutes ago

  @falconer_ai great ship. Most companies are not as "avant-garde" as #coinbase in having on-technical teams shipping production code, at Falconer we strong believe non-tech teams like customer support need all the best AI-support they can get to get things done.

• 
  Catnaps (@CatNapa74) reported 19 minutes ago

  @brian_armstrong He's right though, its changing. Coinbase used to piss me off a lot. But they have flipped the issues. Coinbase is probably going to be a real bank. Jobs may be lost, yeah that does suck. but look at the incentives that are being pffered in place now. Its going to help people.

• 
  Saira Afeefa (@Crypto_Queenie) reported 20 minutes ago

  @malicious_moon @MetalDollarXMD Yep, Mtl Migrated To Our Own L2 (Like Base On Optimism) And Coinbase Skipped The Airdrop Support, Happens With Upgrades. Meanwhile $Xmd Is Pegged 1:1 To A Basket Of Battle tested Stables ($Usdc, $Pyusd & $Usdp), Fully Redeemable, And Built For Compliant On Chain Finance That Actually Gets Used. Different Chain, Different Product, Way Stronger Utility In 2026. Builders Move Forward, Fud Stays On The Old Ticker!..

• 
  Richard Stubbs (@richardstubbs) reported 20 minutes ago

  GitLab's "Act 2": layoffs, flatter management, micro R&D teams, agents running internal work. Same org-compression playbook Dorsey ran at Block, Chesky at Airbnb, Lütke at Shopify. Coinbase, Cloudflare ... You get the picture.

• 
  WallStreet Veteran 90% (@pham_ba66053) reported 31 minutes ago

  My May Advice: $COIN (Coinbase) — Strong buy $MSTR (MicroStrategy) — Strong buy $SQ (Block) — Strong buy $PYPL (PayPal) — Buy $SOFI (SoFi) — Buy $ROKU (Roku) — Buy $U (Unity Software) — Buy $HOOD (Robinhood) — Don’t buy $AFRM (Affirm) — Don’t buy

• 
  karisayswen (@karisaysw) reported 34 minutes ago

  we had a good run. 🙏 @legendapp raised $15m from a16z + Coinbase Ventures and is now shutting down. > the lesson for all defi builders: forget about redpilling users on crypto/onchain, just give them "better yield, faster payments, more control over their money." attaching the full paragrah here: "Mainstream users don't care if a product is onchain or not. They want outcomes. Better yield, faster payments, more control over their money. The product that wins isn't the one that explains crypto better, it's the one that hides it completely. The benefits are felt, not explained."

• 
  iBrokerFreight (@iBrokerFreight_) reported 36 minutes ago

  @coinbase please help

• 
  dy.troll (@dytrollvscx) reported 38 minutes ago

  @fukupapers @coinbase Problem?

• 
  Brian Cubellis (@BackslashBTC) reported 40 minutes ago

  @TR401 @OnrampBitcoin @MTanguma Hey Tyrone, won't speak for Michael but in my view MIC is structurally superior to any single custodian because it creates fault tolerance and redundancy. Zero assets are lost if one custodian has an issue. No single custodian alone can claim that. And it's less about Coinbase specifically, but the reality that any single custodian will suffer from that lack of redundancy at the entity level. The architecture of MIC distributes trust via multisig where no single participant is able to unilaterally move or lose funds. That's the core advantage. And the fact that Coinbase is the largest custodian could be looked at as a negative in that the larger they get, the larger the target on their back from a potential attack perspective. Our fundamental view is that no single entity is infallible on a long enough time horizon, so distributing trust/ownership at the protocol level is a superior design surface.

• 
  No Bu (@NoBu33170345) reported 41 minutes ago

  @CastilloTrading remember when coinbase was down 5 hours and all the alt coins pumped

• 
  PropW (@PropWGlobal) reported 42 minutes ago

  @BlockBeats_News, May 12 — Analyst Murphy released an analysis highlighting short-term bearish signals for Bitcoin. Currently, the #CVD (Cumulative Volume Delta) on @binance perpetual futures has dropped sharply from its April highs and has now fallen below the 90-day median. This suggests that bullish traders are no longer willing to aggressively chase higher prices, while bears have started actively selling and opening short positions. Meanwhile, spot CVD has also begun to “peak and turn downward.” Historically, when this indicator reverses lower, #Bitcoin prices often follow with a pullback — similar to what occurred in November 2024 and April 2025. However, the key difference in the current market compared to previous periods is that Bitcoin’s CVD on @coinbase remains firmly above the 90-day median line. At the same time, the 30-day net inflows into spot Bitcoin ETFs have not shown any significant decline, indicating that U.S. market capital is still providing support. The analyst noted that shorting Bitcoin in the near term remains a high-risk trade. Bears still need to wait for short-side sentiment on #Binance to gain clear dominance, as well as for spot buying demand from U.S. investors to gradually weaken and exit the market.

• 
  ICAR (@ICAR_Global) reported 45 minutes ago

  @AnreiiUzun @coinbase Most times exchanges will not act on your request to freeze user funds unless there is sufficient proof of fraud and a court order effecting that. If you have the scammers wallet details and every other information, lets help you compile a strong report. It is Free

• 
  Vlad 🍩 (@v_lugovsky) reported 45 minutes ago

  AI is now getting blamed for everything almost by default. Outage -> AI's fault Layoffs -> AI's fault A product gets worse -> AI's fault The site is slow -> you guessed it, probably AI too After the Coinbase outage a few days ago, I wanted to check the data and see whether AI is really becoming a new outage cause. Short answer: not in the way people think.

• 
  rep1c.eth (@rep1cxyz) reported 48 minutes ago

  "Stocks always rise after layoffs" - really? Cloudflare: beat earnings, cut 1,100 people → stock dropped 18% Coinbase: cut 700, called it "AI-native" → surprise loss, stock down 5% Upwork: cut 24% of workforce → stock cratered 19% I went through the actual numbers. The full breakdown ↓

• 
  Daniel Whitmore (@micaellainnoce1) reported 53 minutes ago

  @CartelBrother @Vrg8484 @CartelBrother Coinbase Wallet connection issues during migration are usually caused by network mismatch, wallet session conflicts, or claimpad congestion not loss of funds so if you’re stuck connecting, reply or DM with the exact error and it can be diagnosed directly.

• 
  Practically Perfect (@practiklyperfct) reported 57 minutes ago

  @AdamDraper @coinbase @RobinHoodNYC I saw a new layer of security that leap frogs and helps solve the NSA Google announced quantum key problem. I think all three of you would be interested. How can I make intro?

• 
  authmaan▫️ (@auth_maan) reported 58 minutes ago

  i don’t submit kyc because i want to, i submit it because i have to every time i open a new exchange/app; bybit, coinbase, binance, bitget,.. there’s a moment i dislike, the kyc screen uploading my id, my passport, taking a selfie, confirming my address. it all feels smooth and reassuring but somewhere behind the submit button is a system controlled by people i’ve never met, receiving documents that proves who i am, where i’m from, and what i look like. i don’t do it because i trust it. i do it because there’s no alternative every time i hit submit, something feels like it’s leaving my control. not ordinary data, my identity. handed to a company, probably accessible to their employees, stored on servers i’ll never see. that feeling isn’t paranoia. coinbase proved it’s rational, users had their government IDs photographed by a support agent and sold the breach didn’t happen because coinbase was careless, it happened because their agents had to see the documents to do their jobs that’s the architecture and that’s exactly what actually bothers me, not kyc itself because verifying identity before accessing financial systems makes sense what feels off, is that to confirm i am who i say i am, i have to hand over everything that proves it. my full name, date of birth, nationality, my documents.. the verification and the exposure are treated as the same thing, but they shouldn’t be that’s where @Arcium comes in; when kyc runs on arcium, my documents are never processed as readable data. they’re split into encrypted shares; fragments distributed across independent nodes in arxOS. no single node holds my passport no single node knows what the shares represent each one sees noise the verification runs inside an MXE, a multi party execution environment, computing across encrypted shares without ever reconstructing the original document back together the result we get is binary: verified or not no one saw my passport, no one stored my face, the support agent sees a checkmark, not a photograph i live in a part of the world where submitting identity documents to financial platforms carries a different kind of weight, not normal privacy concerns, they’re questions about who has access to information about you, what they might do with it, and there is usually no clear answer every time i complete kyc, i accept a risk i never agreed to, i accept it because the alternative is not using the app anymore and that’s not a fair trade what arcium makes possible isn’t a better privacy policy or a more trustworthy company. it’s a system where the architecture doesn’t require me to trust anyone with my privacy the verification happens my data stays mine because seeing it was not necessary that’s personally the most important use case to me know your customer without actually knowing your customer

• 
  YAROSSLAV (@undefinedKi) reported 1 hour ago

  Everyone was waiting for Polymarket airdrop and a token. We're getting stock on Nasdaq instead Polymarket officially confirmed the IPO on Nasdaq. December 30, 2026 But don't write it off yet. Stock and token are different instruments. Coinbase is public on Nasdaq and still operates inside crypto. A token isn't impossible here Wall Street gets the stock first. Doesn't mean crypto gets nothing later Pray for $POLY

• 
  Grouchy (@GrouchyCrypto) reported 1 hour ago

  @WatcherGuru @brian_armstrong Didn’t he just have coinbase down for over 5 hours?

• 
  Rohit Chauhan (@degenrsc) reported 1 hour ago

  Coinbase ($COIN): The Bet on Agentic Rails Are Live Thesis Sell-side has written the same Coinbase note for three years: strong in bull markets, exposed in bear ones, best treated as a trading instrument not a compounding business. Hold it when crypto runs. Trim it when it doesn't. That framing is now factually wrong. Q1 2026 isn't a strong quarter dressed up in narrative. It's the first quarter where Coinbase's infrastructure thesis moved from management commentary to audited numbers. 99% of all onchain agentic commerce settled in USDC. 90% of it ran on Base. x402 — Coinbase's machine-to-machine payment protocol — processed over $100M in its first live quarter. Amazon announced a Bedrock integration on the same day as earnings. This is not a roadmap. This is a market that already exists, and Coinbase is the settlement layer. What Q1 Actually Shows > Total revenue: $1.4B. Transaction revenue: $756M. Subscription and services: $584M. GAAP net loss: $394M. Adjusted EBITDA: positive $303M. Cash on hand: $10B+. > The top-line miss on GAAP is accounting noise — the gap between -$394M reported and +$303M adjusted is driven by fair value marks on crypto assets held on the balance sheet, not operational deterioration. Watch EBITDA and cash. Both are fine. > The real story is in three new product lines that didn't exist 18 months ago and are already material: > Derivatives: $200M+ annualized in Q1. New all-time high. CFO Haas flagged this as a "big growth driver in 2026" on the Q4 2025 call. It already is — one quarter in. > Prediction markets: Launched January 2026. $100M annualized run rate by March. Two months to nine figures. That's the fastest new product ramp in Coinbase's history. > Coinbase One: Crossed 1M paid subscribers. A recurring subscription business compounding inside the exchange, decoupled from trading volumes. Three new lines. All newly material. None properly weighted in the consensus model. The Infrastructure Thesis, In Numbers > The Amazon Bedrock partnership is the most important announcement this quarter and it received the least analytical attention. > Amazon didn't partner with a crypto exchange. Amazon partnered with payment rails. > The integration: Bedrock developers can now provision AI agents with USDC wallets via Coinbase's developer platform and execute x402 micropayments natively. Every enterprise building on Bedrock — and at this point that's most of Fortune 500 AI infrastructure — has a one-click path to Coinbase-settled agentic payments. > x402 makes this concrete. $100M processed in Q1 across machine-to-machine payments: API calls, agent task completions, data purchases, compute settlements. That number looks small in isolation. It is not. The technical question — does the protocol work, do the wallets provision, does the USDC clear — is answered. It does. They do. It does. > 90% of it ran on Base. Coinbase's L2 is not a side project. It is the execution layer for agentic commerce, confirmed by Q1 actuals. The standing bear case on the AI thesis has always been timing: that agent-to-agent commerce at revenue-relevant scale is a 2028 story. Q1 2026 is the data point that ends that argument. The question has moved from "will this happen" to "how fast does it ramp." The Layoffs Are the Tell 14% of staff, framed by Armstrong as an AI-native transition. Sell-side read this as a cost-cutting signal during a soft quarter. That's the wrong read. Coinbase isn't cutting because the business is contracting. They're cutting because AI is now doing the work that headcount used to do — and converting those savings into margin expansion and infrastructure investment. You don't cut 14% of staff while sitting on $10B cash and posting positive adjusted EBITDA if you're in retreat. You do it if you're rebuilding the org structure for the next phase. The layoffs and the Amazon partnership announced in the same quarter are the same signal: Coinbase is becoming an AI-native infrastructure company, and it's doing it from a position of financial strength. Risks S&S guidance for Q2 is $565–645M, bracketing Q1's $584M — essentially flat. That's the flag. If x402 and Base sequencer revenue are inflecting, it should show in S&S. If Q2 prints at the low end, the agentic commerce monetization timeline is longer than the narrative implies. This is the single data point to watch on the next print. Transaction revenue is still a crypto beta exposure. A sustained H2 2026 bear market compresses the headline number regardless of infrastructure execution. The data center incident is a one-event risk today. In Q1, a technical issue at a data center briefly halted trading on the platform. Coinbase's institutional custody business — their highest-margin, most defensible asset — runs on operational credibility. One event is forgivable. Two is an institutional AUM problem. Entry Framework Current price: ~$201 (weekly close). Not a screaming entry — price is sitting at a macro decision point, not a support level. Key levels: - $185–200: Current decision zone. Held as support this week. Losing this on a weekly close puts $165 back in play fast. - $165–175: First high-conviction entry zone. March 2026 low. Thesis-aligned add zone — fundamentals intact, price at structure. - $145: Strong structural support. Multiple touches across 2022–2023. Only relevant if macro deteriorates significantly. - $250–260: First meaningful resistance above. Breakout above this confirms the infrastructure repricing thesis is being recognized by the market. Sizing logic: Partial position at current levels ($195–205) is defensible if you believe the Q2 S&S print will be the catalyst. Full conviction sizing on a retest of $165–175. Do not chase above $215 — risk/reward compresses significantly at that level. Catalyst to watch: Q2 2026 earnings (August). If S&S clears $620M, the thesis upgrades to HIGH conviction and the $250 target becomes the base case. The Framework The consensus is pricing a volatile exchange with some recurring revenue bolted on. The correct frame is a payment infrastructure company with a volatile exchange attached. The infrastructure layer — USDC stablecoin fees, Coinbase One, custody, derivatives, x402 — now generates roughly $2.3B annualized with structurally limited cycle correlation. The exchange layer adds upside when crypto markets are active. The infrastructure layer grows regardless. Q1 2026 is the quarter that distinction became verifiable. CONVICTION: MEDIUM — Upgrades to HIGH if Q2 2026 S&S clears $620M (confirming x402 and agentic commerce monetization is converting into reported revenue) OR additional enterprise integrations follow Amazon Bedrock. Stays MEDIUM if Q2 S&S prints at guidance floor — the thesis is structurally right but the revenue ramp timeline is longer than Q1 implies. Size for asymmetry, not full conviction.