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:

• 
  tontoon🇺🇸 (@tontoon81) reported 3 minutes ago

  @VulcanForged Take note: Everyday I log into coinbase for the past 3-4 years, I see my account DOWN. Why? Because over 50% of my holdings on coinbase are invested in PYR! More news, More lose!! Down $200,000 but lets keep "forging on!" Forging on to what? I dont even ******* know. But I am here since 2022 buying and holding.....

• 
  Vortex | BIP448 (@theonevortex) reported 3 minutes ago

  @Arthur_van_Pelt Your latest article was once again easily debunked since you clearly have only an extremely basic understanding of how bitcoin works: > Vortex claims above “[BIP]110 does not stop or slow down any spam”. This is such a disgusting and blatant lie that the data spam apologist camp keeps repeating over and over again. This is 100% false, literally every single spam app has updated their software to be 110 compliant. > BIP110 works as a charm and that is proven on at least 2 places. Not even a little bit accurate. You cited two sources as proof and they contradict each other on the exact claim you're making. Source 1 claims zero false positives, 100% of Rule 7 catches were inscriptions, "zero collateral damage." Source 2 says 81.1% of blocked transactions classify as spam. That means by your own second source roughly 19% of what BIP-110 blocks does NOT classify as spam, and across 1.2 billion transactions that is an enormous absolute number of non-spam transactions caught. A 0% collateral claim and a ~19% collateral claim aren't corroboration, they're a methodological dispute you mistook for consensus. The block weight numbers disagree too. Source 1 says 36% of block space filtered, Source 2 says 4.93% of post-inscription block weight. And note what 4.93% actually means: the transactions BIP-110 targets are under 5% of block weight. That's the emergency justifying Bitcoin's first consensus downgrade. More fundamentally, both are backtests. They replay transactions crafted before the rules existed and count matches. Of course a filter catches the patterns it was written against. A BIP-110-compliant data embedding transaction already exists, verified on regtest against the BIP-110 implementation itself with a published reproducible script, and the compliant version came out LARGER with the data pushed into outputs nodes store in the UTXO set forever instead of prunable OP_RETURN. That is what "does not stop or slow down spam" means. The data still goes in, bigger and less prunable. Your backtests don't touch that argument and your sources can't even agree with each other. > Soft forks like BIP110 historically always involved lots of debating, Yes and they also involve lots of testing which 110 has not gone through, not only that but they already had to request the help of several core developers on several occasions due to the incompetence of dathom who essentially vibe coded the entire client. > Bitcoin Core’s highly controversial data spam-friendly direction with their October 2025 v30 release. Bitcoin core didn't take a spam-friendly direction, this is a lie, they simply changed a relay policy filter default that anyone can configure on their own node. > And this poor lad called Vortex apparently still hasn’t come across my previous article “Stop Being Retarded, Bitcoiners, Eat A Bag Of *****”. Not only did I read the terrible writing I replied BOTH on twitter and medium with the only comment on the page, you can read right? > To add, the as we speak still low initial miner signaling (~1% in recent days, against ~15% full nodes signaling) shows that BIP110 is not steamrolling the Bitcoin network Yes because literally nobody is signaling for this terrible software, it's still less than 1%, it literally never even hit 1% signaling yet despite activation only being weeks away. > This is, of course, backwards. No actually, You're equivocating on "centralize" and the argument dies without it. Rich users bidding up an open fee auction is not market centralization, anyone can outbid them, that's the neutral market working. Centralization means template construction, fee capture, and transaction access concentrating in fewer hands. Filters demonstrably cause that: filtered demand routes out-of-band to direct miner submission (Slipstream exists for exactly this reason), which advantages large miners and starves small pools on public relay. Pulling that flow back into the public mempool was the stated rationale for Core v30's OP_RETURN change. Cause and effect on record, opposite direction to yours. BIP-110 adds a second vector: 55% miner signaling makes miners the arbiters of which paying demand counts as legitimate. Consensus-level legitimacy gatekeeping of demand is blockspace market centralization by definition. Also, OP_RETURN is prunable. The filter's demonstrated displacement effect pushes data into UTXO set outputs your node stores forever, ~95% more permanent burden for ~11% more spammer fees. Your node runner argument is an argument against the filter. > BIP110 restores the balance that existed up till early 2023 by enforcing tighter limits on new UTXOs/outputs (scriptPubKeys ≤34 bytes except OP_RETURN ≤83 bytes; data pushes/witness items ≤256 bytes) while grandfathering pre-activation UTXOs 100% false. Point to the block height where a 34-byte scriptPubKey cap or a 256-byte witness element limit was consensus. You can't, because they never existed. Consensus allows 10,000-byte scripts and 520-byte elements, and OP_RETURN size was always relay policy. "Restores the balance" means inventing rules Bitcoin never had and calling it restoration. What existed pre-2023 was an absence of demand, not a ruleset. Your own historical precedent is the strongest evidence against you. 2014: Core cut OP_RETURN standardness to 40 bytes to discourage embedding, Counterparty moved its data into bare multisig, UTXO set bloat got worse, and OP_RETURN was kept as harm reduction because devs concluded you can only steer embedding, not stop it. Same displacement the BIP-110-compliant demo just showed again a decade later. Satoshi never once used consensus rules to filter content. Relay defaults and persuasion, that's it, from the guy who put the Times headline in the genesis block. You're citing him for a mechanism he never touched. And "subsidised storage" is backwards: the witness discount prices data by its real node cost, output data burdens the UTXO set forever while witness data doesn't. Inscribers pay full market rate under correct pricing. Your rules push data back into the expensive place. > On piling multiple systems/data into one chain (BitDNS discussion, December 10, 2010 Real quotes, wrong conclusion. All three show Satoshi declining to add features or telling other systems to use separate chains. Zero of them show him forbidding data via consensus, and your own third quote contains his actual remedy: "a separate system parallel to the bitcoin network." Build elsewhere is not make-invalid. Here's the fact that settles it: Satoshi had unilateral consensus power and used it, disabling opcodes in 2010 over security bugs, during the same period as your quotes. He never aimed it at data. He also documented his anti-spam mechanism explicitly: transaction fees as the flooding defense. Satoshi's answer to unwanted transactions was pricing, not filtering. You're quoting the man whose revealed preferences support my argument. And the BitDNS thread you cite ends with Satoshi proposing merged mining, which embeds other chains' commitments into Bitcoin's coinbase. His solution to "don't pile systems into one chain" was a form of data embedding. Your quote 2 describes users limiting the SIZE of the chain, a uniform limit on everyone. That maps to a block weight reduction soft fork, not to a filter that keeps blocks the same size while selecting allowed byte patterns. Stand on Satoshi consistently or don't stand on him at all. > LN-Symmetry (Eltoo) and related covenant improvements are advanced Layer 2 features. OP_IF in a tapscript is a spending condition, not "arbitrary data." Vaults, BitVM and ln-symmetry ARE the on-chain enforcement, and Lightning only keeps data off-chain because those scripts stay valid, so reclassifying money logic as data to excuse filtering it just proves the filter can't tell the difference. And "temporary" has no precedent because there has never been an expiring consensus rule in Bitcoin's history, so name one. The reassessment clause is circular since legitimate contracts can't "emerge" while consensus-invalid, and a rule renewable yearly by 55% signaling isn't a pause, it's an indefinite freeze with annual review that nobody funds multi-year R&D under. > This is similar to the previous point: we are talking about temporary restrictions on certain tapscript patterns (e.g., OP_IF/NOTIF in executed contexts and large structures). Your sources don't even agree here: one claims zero false positives precisely because no legitimate tapscript OP_IF contracts exist on mainnet, and now wallets "may need updates," both can't be true. And "may need updates" is what breakage is called by the side that caused it, because when consensus invalidates scripts that working software emits that's the definition of breaking compatibility. The tapleaf workaround isn't free either since flattening mixed threshold policies into separate leaves blows up leaf count, depth and control block size, which BIP110 also limits by your own account, so one rule pushes wallets into the structure another rule punishes. Grandfathering only covers coins already on-chain, not vault and recovery protocols whose pre-signed continuation transactions create new outputs, and since deleting keys after pre-signing IS the vault security model those setups can't adapt. Consensus changes ship with tooling and test vectors settled up front like Taproot did, not with "guidance will be provided." > First of all, neutrality doesn’t mean "anything goes forever in any shape or form". Point to the protocol field that marks a transaction "native." There isn't one, everything BIP110 filters is consensus-valid and fee-paying, so "native" means "transactions I approve of" and your no-censorship claim is circular. Your own source says only 81.1% of blocked transactions classify as spam, so "no native transaction is ever touched" is refuted by your own data. Bitcoin's limits have always been uniform, hitting every transaction identically regardless of content, and you conceded the difference yourself: "optional policy" to consensus. Optional was the neutral part. > No, not necessarily. And even if a split is happening, it will be shortlived, not supported by anyone and it is to be expected that the minority split will be quickly abandoned. SegWit required 95% and Taproot 90% because soft fork safety needs an enforcing majority overwhelming enough to rapidly orphan non-compliant blocks, and signaling isn't enforcement, miners have false-signaled before. At 55% you get persistent reorgs with old nodes flip-flopping between chains, and 2017 teaches the opposite lesson anyway since BIP148 was split-courting brinkmanship bailed out by 90%+ hashrate compliance. Now apply your own numbers: 1% miners, 15% nodes, and a mandatory window where BIP110 nodes reject every non-signaling block, meaning at current support they orphan ~99% of blocks and follow a ~1% hashrate chain with multi-hour block times, rejected by the economic majority. Your "shortlived minority split, supported by no one, quickly abandoned" is an accurate forecast of your own chain, and nobody needs to request exchange listings because a persistent chain with divergent rules is a separate coin by definition. > No. Again an argument where Vortex has it backwards. The actual attack has been ignited by Bitcoin Core. Your threshold defense is a confession. High thresholds don't "reflect less contention," they exist BECAUSE contention is dangerous, the threshold is the safety margin against a persistent split, and contention is exactly when the other 45% of hashrate really will keep mining the old rules. So "higher thresholds stall needed fixes" translates to "the safety mechanism keeps stopping us, so we lowered it," making BIP110 the first activation design in Bitcoin's history built to defeat contention rather than require its absence. And 55% of what, miner signaling? Bare miner majority deciding consensus rules is the exact model the 2017 UASF movement existed to reject, users decide rules and miners order transactions, now inverted to "55% of hashrate suffices over user objection" at 15% nodes and 1% miners by your own count. The Luke and Szabo screenshots criticize a relay default anyone can override today and contain nothing about activation safety, so point 7 stands unanswered, appendix included. > BIP110 doesn’t claim to eliminate all spam forever as spammers can indeed fragment data, use other attack vectors, or wait for expiry of the BIP. Your point 1 called "BIP110 does not stop or slow down spam" a disgusting and blatant lie, and your point 8 opens by agreeing spammers "can indeed fragment data, use other attack vectors, or wait for expiry." That was the claim, you've ended your rebuttal by conceding the sentence you began it calling a lie. "Raises the bar" doesn't survive the math either, ~11% more spammer cost for ~95% more permanent node burden with the compliant variant demonstrably larger than the original. The "already working" tweet loses both ways since inscription decline is a market-wide cycle collapse no 1%-signaling proposal gets credit for, and spammers "making amendments" IS the adaptation thesis, the failure mode cited as proof of success. The 85% of nodes running clients without BIP110 already expressed their position in software, silence isn't mandate, and seven years of free BSV blockspace with zero migration settles "they'll just leave." Notice the pattern: activation is success, failure strengthens the protocol, spam down proves it works, spam adapting proves maintenance works. A theory every outcome confirms isn't engineering, it's faith. Summary: See insults don't help your case, neither does your lack of basic knowledge of how bitcoin works. Here's the scorecard, sources that disagree with each other cited as dual proof. A "restoration" of consensus rules that never existed. Satoshi quoted for a mechanism he never once used while his actual anti-spam tool, fees, argues my side. Breakage conceded as "updates," neutrality conceded as "optional to consensus," split risk conceded as "designed for contention," and the original claim you called a disgusting lie conceded verbatim in your own closing point. Every substantive argument you made either contradicted your sources, contradicted your other arguments, or contradicted itself. I almost didn't even need to debunk this because this entire post is basically a confession of having no idea what's going on.

• 
  v1nce.base.eth (@v1nce_base) reported 5 minutes ago

  @MLeeJr Or coinbase should have tried to support their builders

• 
  Sane Stock Weekly (@SaneStockWeekly) reported 6 minutes ago

  @scottmelker 50+ days of negative Coinbase premium = U.S. spot buyers on sidelines. When institutional demand is this suppressed while price holds, either global demand is carrying it — or the next leg down has no floor. Either way, this metric deserves a close watch right now.

• 
  Eezzy 🧸 (@notEezzy) reported 7 minutes ago

  @crypto_bitlord7 Monad scammed you with a Presale in partnership with Coinbase Then gifted their CT faves a 5 figs airdrop for doing nothing Those CT faves dumped their airdrops on dumb retail that bought the presale Robinhood chain is nothing like that **** And Robinhood chain will save us all.

• 
  𝕯𝖊𝖑𝖊 (@dele_dave) reported 8 minutes ago

  On June 30, some of the biggest companies in payments lowkey challenged Circle & Tether. Visa. Mastercard. American Express. Stripe. BlackRock. Coinbase. Google. Shopify. Along with more than 140 partners, they launched Open USD here @openstandard. The interesting thing here is the business model. Stablecoin issuers make money by holding users' dollars in short-term Treasuries and keeping the yield. It's one of the most profitable businesses in finance. But there's always been a catch. Issuers earn most of the economics while distribution partners bring in the users. Circle already shares a large portion of its revenue with Coinbase because distribution is that valuable. But OpenUSD model is diff. Instead of one company keeping the yield, the economics are shared across the partners that distribute the stablecoin. In other words, distribution decided it no longer wanted to be just a customer. It wanted to become the issuer. That's why Circle's stock fell 18 percent after the announcement.

• 
  5 (@w5rldwide) reported 10 minutes ago

  Anybody else having issues with Coinbase transfers ?

• 
  James carver (@jamesc559) reported 12 minutes ago

  @coinbase Pretty hard to move your crypto when you don’t know when your system is down because you don’t let people know before they make purchases

• 
  blockhopper256 (@blockhopper256) reported 14 minutes ago

  @blockchainchick Therefor the plan is that once the coinbase reduces slowly transaction volume or demand to be positioned on the block increases which increases the fee. Therefor by 2140 we should be in a structure where fee’s cover the value or cost of securing the network with energy.

• 
  C O L E E N ♡ 彡 (@coolsgp19) reported 15 minutes ago

  @coinbase Help me understand specifically why its over a month now and still no update “we will update you” is no longer acceptable. Can I access my account now?Please?😫 @CoinbaseSupport

• 
  landö (@landosworldd) reported 15 minutes ago

  @coinbase What ******** is taking so damn long

• 
  JohnnyCryptoseed (@CryptoseedJ) reported 15 minutes ago

  COINBASE Gouges lower tier users - 1.9% per limit order settled on COINBASE WTF! They put a 5 day hold on your flow too. It is getting HARDER to ONBOARD for free. BANK inlfuence is everywhere! **** BANKS!

• 
  bizzareX (@x_bizzare) reported 15 minutes ago

  @RambleGG @coinbase @CoinbaseSupport How about you fix yourself and stop using them !? After all this time im starting to believe you guys are doing paid fud why else you keep using them ?

• 
  Inux (@0xInux) reported 19 minutes ago

  the lever isn't the $1m dev fund - that's a rounding error. it's a $100b profitable parent that can route $307b in customer assets into its own dexs, as the issuer taking the spread. base/coinbase distribution model, not a mercenary L2.

• 
  StFixe (@StFixee) reported 23 minutes ago

  @MrKrabsOnChain i'm not even in cuz coinbase is slow asf rn