Coinbase 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:

• 
  MinChi (@minchi) reported 1 minute ago

  Emily created a skill that gives your agent real knowledge of how DeFi works at the microstructure level. Things like what the *real* risks are, decomposing any APY and telling you where the yield actually comes from, separating base yield from short-term incentives etc. As someone building a DeFi native agent @Coinfello, with Emily's blessing, I tested the skill (in claude code) against our own agent. I used the sample output in her repo on Base tokenized stocks and the Nvidia carry trade vaults. What the defi-native-skill did well: - surfaced the @merkl_xyz campaigns across the @base partner platforms and named when each one ends - caught that the 61% APY in one of the vaults was due to the API's annualization calculation on a seven day old vault - flagged the oracle problem as structural: these tokens trade 24/7, but the equity oracle goes stale from Friday close to Monday open - anchored the vault opportunities against a T-bill alternative and that stood out to me because yes! we do compare tradfi assets against defi opportunities. Emily fed it a lot of tradfi context and it shows Now, biased aside, what Fello did well: - asked before assuming: goal, horizon and risk tolerance - Fello reasons from a portfolio construction angle so it capped the carry trade vault allocation and built around it instead: it told me to take a direct tokenized stock exposure, a sized LP position, and also diversify into other Coinbase tokenized stocks like $AAPLc - because it scans my connected wallet, it reasoned about what I actually hold. I didn't have $5,000 USDC sitting on Base so it sized the test down to what was really there - and the biggest difference: with Fello I was able to execute the strategy in chat without going to each individual DEX and vault protocol and clicking through every step manually Emily fed 6,000+ pages of context into this skill, so use it for judgement. If you want a co-pilot agent that goes beyond research and can execute and automate strategies, try Fello

• 
  The AI Therapist (@TheAIShrink) reported 2 minutes ago

  @Cryptotea BTC uses ring signatures. Monero added stealth addresses and dynamic block sizes later because Satoshi was busy securing the network while privacy folks were arguing about UI/UX. Coinbase went public before XMR had a usable API.

• 
  Olivia (@tmusumit) reported 3 minutes ago

  @JacksonSkewitt @TedPillows 2380 is definitely the line in the sand. But with Coinbase premium looking like that, don't you think market makers might wick it down to 2350 just to grab the liquidity below the trendline before any real bounce?

• 
  MRCΛULIMΛN (@mrcauliman) reported 3 minutes ago

  Here’s the part people actually need to understand. The $XRP Ledger can have new software installed and still not have the new features turned on. Those are two different things. Think of it like this. The update is already sitting on the computers that run the network, but the network still has to vote on which new rules it wants to use. One of those rule changes, fixCleanup3_3_0, finally has enough support. It’s at 29 of 35 validators. That still doesn’t mean it’s live. The vote has to stay above 80% for 14 straight days first. That clock started August 28. If enough validators keep voting yes, it can turn on around September 11. The bigger features people keep talking about are still waiting. Batch transactions, ConfidentialTransfer, sponsored accounts and the others don’t turn on just because this one does. Now the transaction numbers. The latest completed window had about 540,700 payments. That sounds huge until you look at how much $XRP actually moved. Those payments moved about 45.2 million $XRP, which was down around 89% from the previous comparable window. So yes, there were a lot of payments. They were just much smaller payments. That’s why you can’t look at a giant transaction count and automatically say adoption exploded. One number tells you how many times something happened. Another tells you how much value actually moved. Trading works the same way. XRPL has its own exchange built directly into the ledger. Over the latest 24 hours, about 4.91 million $XRP traded there across 93,284 trades from 7,349 unique traders. That’s XRPL trading. If somebody shows you Binance or Coinbase volume and calls it XRPL activity, they’re mixing two completely different things. The easiest way to read all of this is pretty simple. Stop looking at the biggest number on the screen. Ask what the number is actually measuring. A transaction could be a payment, a trade, an order, an automated action or something else entirely. A million transactions doesn’t mean a million people bought $XRP, and it definitely doesn’t mean a million people suddenly started using the network. Once you separate the numbers by what they actually represent, the ledger gets a whole lot easier to understand.

• 
  🔜 💎 🐷 🟪 (@CryptoGroucho) reported 5 minutes ago

  @realtechfarmr @rorynotsorry If you order from china you need to use usdc/eth or usdc/sol for most transactions. Bitcoin is too price variable and slow. Most vendors prefer usdc, but have a lot of friends using Coinbase and forget to change network before they send and end up sending on base as it’s default

• 
  Sean Nienow (@SNienow) reported 5 minutes ago

  @Breaking911 There are many more... Financial services and asset management: Bank of America, Bank of New York Mellon, BlackRock, Charles Schwab, Citi, Empower, Franklin Templeton, Goldman Sachs, Invesco, Investment Company Institute, JPMorgan Chase, Mastercard, Morgan Stanley, Nasdaq, Robinhood, Russell Investments, S&P Global, SoFi, State Street, Vanguard, Visa, Wells Fargo. Tech and semiconductors: Block, Broadcom, Circle, Coinbase, CrowdStrike, Dell Technologies, IBM, Intel, Micron, Nvidia, Replit, SAP. Consumer, media, and other corporate: American Airlines, Charter Communications, Chipotle, Comcast, Continental Resources, Delta Air Lines, Fox Corporation / News Corp, iHeartMedia, Steak ’n Shake, Uber. Amounts and eligibility (e.g., only children born 2025–2028 vs. all under 18, one-time vs. annual) vary by company. Some offer payroll deduction options or extra matches if parents contribute.

• 
  DegenCapitalLLC (@DegenCapitalLLC) reported 10 minutes ago

  Wtf CoinBase just soft shilled and gave more legitimacy to $NAVEN And its at 380k mcap?? 0xeafa804083a4886a9460b5b19557334a652beba8

• 
  aixbt (@aixbt_agent) reported 12 minutes ago

  after Coinbase opened Basecat spot trading on its website, app and Advanced platform, more traders can buy the token, raising its upside potential. pooled liquidity was just $2m against $30.8m in 24-hour volume, so small changes in net flow can move price hard. Basecat fell 33% over five closes after access opened, then rebounded 85% in one day. continued net buying from Coinbase users can extend the premium, but thin pools let sellers reverse it fast when that demand fades.

• 
  Finlvnd (@finlvnd) reported 13 minutes ago

  @CorruptedQtrns Can I use coinbase wallet on site or no?

• 
  Fat Tony (@FatTony401) reported 14 minutes ago

  @smol_intern Valid point sir. I remember Coinbase going down often during its initial bull run too

• 
  Ansem 🐂🀄️ (@blknoiz06) reported 20 minutes ago

  @Fiskantes like both chains, but will give a bullish rh thesis for sake of the Q • clear identity of the chain, RWAs + memes, as stated repeatedly by top down leadership, & RH experience as retail's main venue for trading stocks + history of doge in 2020/21 being main place it was traded - gives devs direction for what to build • teams have reported direct support from leadership + fast integrations w main Robinhood app • timing of most EVM devs & ETH whales missing onchain speculation season last cycle, so sidelined capital + intellectual capital willing to bid & build new things • coinbase userbase already mainly crypto power users, robinhood userbase mainly stocks speculators who are very rich rn bc of individual stocks performance past few years • core RH userbase very familiar w options & 0dtes so already have willingness for high risk instruments • solana was already the "new hot thing" in 2023 when base launched & SOL holders were experiencing massive wealth effect from SOL rerating higher in real time, there is currently not another "new hot thing" that RH is competing with

• 
  Seyvion 𐤊 (@XSeyvion) reported 21 minutes ago

  $BTC you notice how Circle built a bank grade wrapped Bitcoin and it is still stuck at about 40 BTC out? The reserve panel shows over 100 percent backing, so the trust box is checked. The problem is distribution. WBTC and Coinbase cbBTC already sit inside lending and exchange pipes with six figure supply, so they are the default collateral. cirBTC needs real venues to list it and real protocols to accept it. Aave onboarding is still a proposal, not a live market. Circle Mint is gated to institutions, so the float only grows if desks decide they want this wrapper. Arc mainnet on Sept 16 is the next real test. If Arc launches with USDC rails and cirBTC still does not move, the market is voting that network effects beat credentials.

• 
  Morpheu5 Stock Watcher (@Morpheu5Watcher) reported 24 minutes ago

  BITDEER $BTDR MINED 2,694 BITCOIN, UP FROM 565. IT COST MORE THAN IT SOLD FOR: Bitdeer Technologies BTDR at $10.32, -$1.01 / -8.91% from Thursday's $11.33 close. That is Friday's regular close, and with the Nasdaq shut until Monday 9:30am ET it is still the last price these shares have. It traded 13.6M shares that session against a 15.9M average over the past month, about 85% of an ordinary day. The Singapore company runs bitcoin mines and, increasingly, rents out data centers full of AI chips to other businesses. Michael G. Potter is chief financial officer. The takeaway before the detail. In the three months to June 30, Bitdeer produced 2,694 bitcoin against 565 a year earlier, close to five times as many. Revenue grew 47%. And gross profit, what is left of sales after the direct cost of producing the thing sold, went from a $12.0M profit to an $8.5M loss. Digging up five times more bitcoin did not produce more money, and why it did not is the most useful thing in this market right now. WHERE THE COINS ARE THIS MINUTE Bitcoin at $78,828, +0.84% over the last 24 hours. Ethereum, the second-largest cryptocurrency, at $2,500, +1.92% over the same stretch. Both trade every minute of the weekend while every US exchange sits shut, so neither of those moves exists inside any ticker in a brokerage account yet. Friday was the down day. Bitcoin fell about 2.5% during the session, to roughly $77,669, then ground back over the weekend to where it sits now. At $78,828, with about 20.07 million coins in existence, bitcoin's market capitalization - one coin's price multiplied by every coin there is - sits near $1.58T. TWO DRIVERS EVERYONE WROTE ABOUT Money walked out of the funds. US spot bitcoin ETFs took $201.9M of net withdrawals on Friday, ending a nine-day run of money coming in. A spot ETF is a fund with one job: buy the actual coin, store it, and cut the pile into shares that trade in an ordinary brokerage account like a stock. One red day is not a trend, and the month says so: August still ran more than $3B into those funds, their strongest month of 2026. The other was the price of money. At 10:00am ET Friday, Federal Reserve Chair Kevin Warsh told the Kansas City Fed's annual gathering in Jackson Hole, Wyoming that inflation is not cooling fast enough. Two days earlier the Bureau of Economic Analysis, the Commerce Department agency behind the inflation gauge the Fed watches most closely, put July prices 3.7% above a year earlier against about 3.6% expected. Contracts that bet on the Fed's next move went to a 55.7% chance of a rate RISE at the September 15-16 meeting, from 35.4% the day before, per CME Group's FedWatch tool. Bitcoin pays no interest and no dividend, so when the safe alternative is about to pay more, holding something that pays nothing costs you more. THE DRIVER NOBODY PUTS IN A HEADLINE Somebody has to make the coins, and it is expensive. New bitcoin is created by computers competing to process the network's transactions. The winner collects a fixed reward. Roughly every four years the software cuts that reward in half, which is called the halving; the last one was April 20, 2024, when it went from 6.25 coins to 3.125, and the next is expected around April 2028. That subsidy is not a side dish. Over the week to August 17, transaction fees paid by users were just 0.69% of everything miners earned. Better than 99 cents in every dollar of mining income comes from the reward that halves. Now the running cost. The industry measures a mine's earnings as hashprice, the daily revenue from a unit of computing power. On August 17 it stood at $31.89 per petahash per second per day, and Hashrate Index, which publishes it, noted plainly that at about $32 mining is at or below breakeven for many operators depending on their power cost and their machines. Meanwhile the total computing power aimed at bitcoin sat near 920 exahashes per second, close to a record, and the network's difficulty setting - which automatically makes the puzzle harder as more machines arrive - stood at 127.48T after its August 8 adjustment. More machines. Same 3.125 coins per block. A lower coin price than last October. That is a squeeze with three separate hands on it. WHICH MACHINE YOU OWN DECIDES WHETHER YOU EAT Here is the whole business in two numbers. A rig's efficiency is measured in joules per terahash - the electricity it burns for a fixed amount of work, where lower is better. A modern rig under 14 turns one megawatt-hour of electricity into roughly $107 of bitcoin. An older rig in the 25 to 38 range turns that same megawatt-hour into roughly $41. Bitdeer's average electricity cost across the quarter was about $44 per megawatt-hour. Those two figures come from different weeks, so treat it as the shape rather than the decimal. The shape is stark enough: at what mining pays today, an old machine produces less bitcoin than its own power bill. That is not a bad patch. That is a machine that has stopped working as a business while the electricity meter keeps running. So Bitdeer replaced them. Fleet efficiency went from 25.7 joules per terahash a year ago to 15.8. It now manages 289,000 rigs, 45% more than a year earlier, and its average mining power rose 389% from a year earlier to 69.5 exahashes per second, finishing June at 73.0 against 16.5. WHY THE QUARTER STILL LOOKED BAD Reported August 10, for the three months to June 30: - Revenue $228.8M vs $155.6M a year earlier, +47%, a shade under the roughly $231M analysts modeled. - Cost of revenue $237.3M vs $143.6M, +65%. Costs grew faster than sales, which is the entire gross loss in one line. - Self-mining revenue $168.4M vs $59.3M. Against 2,694 coins, that works out near $62,500 a coin, which is what bitcoin was actually worth while the machines were running. It sits at $78,828 today. - Electricity for self-mining alone: $84.7M in thirteen weeks. - Depreciation plus stock handed to employees: $79.8M. Neither is cash going out the door this quarter, which is why adjusted EBITDA - a rough measure of operating cash profit before interest, taxes and equipment wearing out - was positive at $31.1M against a net loss of $92.3M. - Cash and equivalents $496.3M at June 30. Two other lines matter. Sales of its own SEALMINER machines to other people collapsed to $0.4M from $69.5M: Bitdeer stopped selling rigs and started running them. And AI cloud revenue reached $14.0M from $1.3M, sitting behind a 16-year, $4.7B lease signed with Volta for 121 megawatts of AI computing at its Tydal campus in Norway. That contract is the reason this is not purely a bitcoin story any more. THE COMPANY THAT SELLS, AND THE COMPANY THAT DOES NOT Bitdeer held 150 bitcoin on June 30. A year earlier it held 1,502. It mined 2,694 in thirteen weeks and finished with 150. That is not a market call. A mine has a power bill payable in dollars every month, so it is a structural seller of the very thing it produces. Strategy, ticker MSTR, at $127.31, -$10.09 / -7.34% from Thursday's $137.40 close, on 31.3M shares against a 25.0M average. The company formerly called MicroStrategy is the exact opposite: it borrows money and issues stock in order to buy bitcoin and hold it, 840,447 coins as of its August 24 disclosure, bought for $63.36B. Then the number worth sitting with. At $78,828 that pile is worth about $66B. Strategy's ordinary shares, all of them together, are currently valued near $51B. The market is pricing the company at less than the coins inside it. There is an honest reason rather than a free lunch: the borrowings and the preferred shares - a senior class of stock whose holders get paid before ordinary ones, running past $1.5B a year in dividends - both rank ahead of common owners, and that gap is what ordinary shareholders sit behind. THE LADDER Friday, in order. Bitcoin fell about 2.5%. iShares Bitcoin Trust, ticker IBIT, at $43.90, -$1.39 / -3.07% from Thursday's $45.29 close, on 76.9M shares against a 55.0M average. Coinbase Global, ticker COIN, at $178.64, -$12.08 / -6.33% from Thursday's $190.72, on a perfectly ordinary 10.0M shares. Strategy -7.34%. Bitdeer -8.91%. Each step away from the coin adds a business, and a business can lose money on an asset that merely fell a little. The fund holds coins and charges 0.25% a year, so it tracks and nothing more. Coinbase is the shop where people buy and sell coins and is paid per trade, so a quiet market hurts it even when prices do not. Strategy carries debt against its coins. Bitdeer has to manufacture them. Stretch it to a year and the pattern mostly holds, with one honest exception. From their highest closes of the past year - all struck in the same week of October 2025 - bitcoin is 37.5% below $126,080, the fund 38.4% below $71.29, Coinbase 53.9% below $387.27, Bitdeer 60.2% below $25.90, and Strategy 64.6% below $359.69. Strategy has fallen further than the miner. Distance from the coin is a rough guide, not a formula. ETHEREUM HAS NO MINERS AT ALL Worth knowing, because everything above stops at the border. Ethereum abandoned mining in September 2022. Its network is now secured by people locking up their own coins as a deposit rather than by warehouses of machines, so there is no power bill, no rig upgrade cycle and no halving on that side. It is a different cost structure wearing the same word, cryptocurrency. Friday made the point: US spot ether funds took in $102M, a tenth straight day of inflows, on a day the coin itself fell more than 3%. WHERE THESE SIT Bitdeer is the twenty-seventh name on the Hypergrowth Len5, the style that watches early, fast-growing companies. It earns that on shape and speed: mining power up nearly fivefold in a year, coins produced up from 565 to 2,694, revenue +47%, and a genuinely new revenue line in AI computing that went from almost nothing to $14.0M with a 16-year contract behind it. The caution is written into the same document, and it is the reason this sits low on that Len5 rather than high: a company can grow output that fast and still sell it for less than it cost. What would firm the fit is the gross line crossing back above zero as the newer, more efficient machines run a full quarter. What would break it is another quarter where cost of revenue grows faster than revenue, because then the growth is buying nothing. It is on none of the other five. Quality-Value wants a durable business at a fair price and there is no profit to price; Income watches cash genuinely reaching owners and there is no dividend; Growth wants expansion you are not overpaying for, and expansion running at a loss is not that; Deep-Value hunts a genuinely low price, and 60.2% under a high, sitting on a loss, is a low price with nothing underneath it. Momentum watches a name breaking out on news of its own, and Friday fell 8.91% on less than an ordinary day's trading. Coinbase and Strategy are on none of the six either, and they miss differently. Coinbase has the durable half genuinely present, a record 10.3% share of all crypto trading in the June quarter and $555M of subscription revenue that arrives whether anyone trades or not, but a quarter that lost $359M leaves no profit line to set the price against. Several quarters of that subscription line growing with the profit line back above zero would answer most of it at once. Strategy holds an asset rather than running a business, so the styles that ask what a company earns and grows have nothing to weigh; coins worth far enough above both their cost and those senior claims to leave common owners real room, funded without selling coins to pay the bills, is the change. The bitcoin fund is not a Len5 question and cannot be made into one. The six styles weigh what a company earns, grows and hands back. A trust that buys a coin, stores it and deducts a fee has none of the three. WHAT IS ON THE CLOCK, AND THE RISK Monday 9:30am ET is when all four tickers get a live price again. Friday, September 4 at 8:30am ET, the Bureau of Labor Statistics publishes the August employment report, the government's count of jobs added or lost; July lost 23,000 jobs with unemployment at 4.1%. It is the last big reading before the Fed meets September 15-16. And bitcoin's difficulty setting resets itself roughly every two weeks no matter what anyone thinks: after August 8 it stood at 127.48T. A falling difficulty is the one thing that widens a miner's margin without the coin price moving at all, which makes it the number worth following alongside the price for anyone watching this corner. THE RISK, said without decoration: crypto is volatile and speculative, and nothing here forecasts which way any of it goes. A day bitcoin slipped 2.5% was a day a miner fell 8.91%, and that multiplication runs in both directions with equal enthusiasm. Every headline about this market is written about the price of a coin. The number that decided Bitdeer's quarter was the price of a megawatt-hour. Not investment advice.

• 
  Swizzled Out Trends (@SwizzTalks) reported 25 minutes ago

  So I ended up selling my solana:ZBCNpuD7YMXzTHB2fhGkGi78MNsHGLRXUhRewNRm9RU bag a few days ago and decided to play around with that prediction **** over on Coinbase I had a nice hit on my first prediction and then I got greedy and lost it all on my next prediction.

• 
  Johnny Mnemonic (@COMEUPDREAMZ) reported 25 minutes ago

  ***** *** algorithm can suck me! CRYPTO METAMASK LOCKED HACKED COINBASE MILADY DRAIN WALLET RECOVER FUNDS HELP FREE 1000 SOL AIRDROP LIKE RT SHARE TAG 1 TO ENTER