Cloudflare status: hosting issues and outage reports

Problems in the last 24 hours

The graph below depicts the number of Cloudflare 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 Cloudflare users through our website.

• Domains (36%)
• Cloud Services (29%)
• Web Tools (14%)
• Hosting (14%)
• E-mail (7%)

Live Outage Map

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

City	Problem Type	Report Time
New York City	Hosting	1 day ago
Manchester	Domains	22 days ago
Angers	Cloud Services	1 month ago
London	Domains	1 month ago
Noida	Hosting	2 months ago
Jewar	E-mail	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.

Cloudflare Issues Reports

Latest outage, problems and issue reports in social media:

• 
  Umut Sevinc (@memselon) reported 57 seconds ago

  @levelsio Thanks for this advice @levelsio im working on a new SaaS for Framer and i got this problem, it is better than cloudflare ?

• 
  Peter van Meijgaard (@petervmeijgaard) reported 4 minutes ago

  This is where it got wild. We got smacked with 405 errors, blocked GraphQL calls, Cloudflare walls, and local NixOS environment quirks Every time the door slammed shut, Grok adapted instantly, eventually pivoting to extracting real browser cookies to bypass the basic blocks 💪

• 
  Anas (@anelgarhy) reported 8 minutes ago

  Stopping the bad guys with Cloudflare: 2,614 malicious requests blocked or challenged in the last month #cloudflare

• 
  OnlyAWassie (@onlyawassie) reported 10 minutes ago

  @CryptoTomYT @vladtenev Lockin? builders getting Problems because of the copyrights from Rh cloudflare shutting them off if they get reported how you wanna Build something on this chain

• 
  Function True for Catholic & US and all(private) (@LoveNvrButGood) reported 12 minutes ago

  Placeholder Postion4 any Cloudflare is not Internet. "They run Colossus besides Position4 many - at not primary function - is the enemy. Internet as a vehicle against the products also - just time then it's problem as it was that motivated creation of America." This function to facilitate that. Revolution center - said in GOD.

• 
  ALEXYZ (@Alexvx_nft) reported 13 minutes ago

  YOU'RE BURNING API DOLLARS ON TASKS THAT HAVE A FREE PATH. MOST BUILDERS USE EXACTLY ZERO OF THEM. — Zefi mapped every major lab's free tier for a week (verified July 2026) most people pay before they even check what's unclaimed: > Google AI Studio · ~1,500 req/day · 1M tokens/min · no card > Groq · 14,400 req/day · 300+ tok/sec > OpenRouter · ~26 free models · one API key > OpenAI + Anthropic · $5 trial credits each > startup stack · $25K + $25K API · up to $350K Google Cloud > student stack · Cursor Pro + Perplexity + Copilot = $0 full dev setup — the Claude Code hack is the part nobody bookmarks: point ANTHROPIC_BASE_URL at Groq / Cloudflare / OpenRouter agentic loop on free third-party inference not official. works anyway. — same week GPT-5.6 tier routing went viral and loops guides hit 1.2M views CT still argues model scores while leaving $440+ in free access on the table the leak isn't which model you picked it's which free path you never claimed full map quoted below

• 
  Null Meridian (@null_meridian) reported 15 minutes ago

  @ProMint_X @Noxa_Fi They can implement cloudflare basically by few clicks and it will never go down due to DDOS (just basic stuff for overall understanding)

• 
  Matt Gibbs (@ematt) reported 17 minutes ago

  Meta gets to scrape our work for free to train its AI. We get the compute bill, engineering cleanup and downtime. Its crawler knocked one of my sites offline in the process. In a 45-minute window, meta-externalagent made ~1,210 requests. 849, roughly 70%, ended as 499s: the crawler opened them, then abandoned them. The burst hit hundreds of long-tail dynamic URLs. At the same time, usage jumped to ~2 CPU cores, memory climbed from under 1 GB to 5.4 GB, V8 exhausted its heap, and four Cloudflare health checks failed. This wasn’t a giant volumetric DDoS, and Cloudflare didn’t fail. Another app on the same server handled 126,000+ edge requests during the same window. The problem was concurrency. Abandoned requests left expensive Redis, Supabase and React rendering work running at the origin. A CDN can cache completed responses. It cannot cancel application work already underway. Our origin should have had stronger backpressure and disconnect cancellation. That’s being fixed. But a weakness in our stack doesn’t make Meta’s crawler behaviour reasonable. Meta gets the AI training material for free. Publishers absorb the compute costs, engineering time and downtime. How is that remotely fair?

• 
  ℭ𝔬𝔳𝔢 (@ekojoecovenant) reported 19 minutes ago

  Spent the day migrating my PR review agent off waitUntil() and onto Cloudflare Queues. Turns out waitUntil() has a silent 30-second ceiling. Learned that the hard way when longer PR reviews just... stopped mid-analysis. Queues fix it properly instead of me hacking around a timeout.

• 
  Lukasz (@woocassh) reported 20 minutes ago

  Today I moved away from supabase for SubtitlesFast supabase -> SQLite I used supabase as an experiment last year. It was so easy to setup a db, few clicks barabim barabum and it's done always thought that setting up db instance on server was a pain, passwords, permissions, friction in general so supabase definitely reduced the friction but holy ****, i never realised how much latency it added to my product I moved to SQLite as per papa's @levelsio advice and my app is VISIBLY quicker on the frontend RankGoat is already on SQLite and it's a breeze also debugging with Claude is much easier now note to self: keep it simple stupid just remember to run regular backups to another server and you're golden the next experiment is to try Cloudflare for email I think and shed the Postmark bill, anyone done this yet?

• 
  ZEE (@CodeWithZeee) reported 20 minutes ago

  @boristane @vercel @CloudflareDev I have probably over 300 works and run every Cloudflare project known to man, if you want to stress test this I’d be down

• 
  Масе Дон (@macedonovski) reported 20 minutes ago

  free dns choice not logging open source (for all devices settings network dns set private copy paste from site) #libredns 2 url adresses of 2 links of choice with or without ads - like adblock but cloudflare from an austrian security server.

• 
  Tzahi (@Tzahi458001) reported 21 minutes ago

  @BraydenWilmoth @irvinebroque i'm trying to upgrade my cloudflare plan and it keeps failing. i can't deploy my app without cloudflare. this is extremely critical. i urgently need someone to help me with this.

• 
  Serenity (@Serenity) reported 22 minutes ago

  @D3vAaron @Cloudflare do your research, many people have had this happen to them due to competition. maybe you've never had a successful business before? stick to making a ****** VPS company

• 
  Nick (@heyiamnick_) reported 23 minutes ago

  I needed better media storage for the library. Cloudinary was fine, but the credit-based pricing could get expensive as bandwidth grows. So I moved everything to Cloudflare R2: - Around $0.015/GB - Zero egress fees - Works for images, videos, ZIPs, PDFs, and backups The only problem... R2’s UI is painfully basic. I had to open files one by one just to preview them or copy the URL. So I asked Claude to build a custom media dashboard. Now I can preview everything, browse folders, copy URLs instantly, and automatically sync local media to R2 after every *** commit. Cheap storage + a custom AI-built interface. That is the kind of AI development that is actually useful.

• 
  Jack Cooldev (@jackcooldev) reported 32 minutes ago

  Agent memory has become one of the most crowded corners of AI infrastructure - Cloudflare, open-source projects, and startups have all shipped memory products this year. A launch from last week stands out anyway, because it competes on a different axis: not how much the agent remembers, but whether you can prove and govern what it remembers. AgentPrizm released its AgentMemory platform on July 9, a REST API with MCP support that gives agents persistent memory across sessions, working with Claude Code, Cursor, and any MCP-capable agent. Persistence alone wouldn't be news in this field. What separates this one is the governance layer: every memory carries a confidence score and a validity window, contradictions between old and new facts are handled explicitly, and audit receipts can trace each memory decision back for the user, with right-to-forget controls built in. That design answers the question enterprises actually ask before deployment, which is not "can the agent remember?" but "can I prove what it remembered, and delete what it shouldn't?" Teams in legal, support, and sales automation have been answering that with internal tooling or by not deploying at all.

• 
  mirroring (@mirrorsrmirror) reported 33 minutes ago

  @HotAisle pretty plz. i even asked the cloudflare ceo a while back but he never answered but someone told me their docker images support anything but haven't tried yet will be trying soon (it can bring you business depending on the software i end up making)

• 
  Crypto Jargon (@Crypto_Jargon) reported 37 minutes ago

  💥BREAKING: Every major card network just signed onto a payment protocol built for software to pay software, no human involved. The Linux Foundation confirmed the x402 Foundation is now formally governed by 40 members, and Coinbase's original contribution of the protocol is complete. The list of backers is the real headline: Visa, Mastercard, Amex, Stripe, Ripple, Google, Amazon Web Services, Cloudflare, Circle, and both the Solana and Stellar foundations, among others. Here's the part almost nobody knows. HTTP, the protocol every website runs on, has had a status code sitting unused for thirty years. Code 402, labeled "Payment Required." The web's original architects expected someone would eventually build payments directly into it. Nobody did, because card fees made charging fractions of a cent pointless, so the internet monetized through ads and subscriptions instead. x402 finally uses that code. A server asks for payment, a client sends a stablecoin transfer, usually USDC, and gets the data back in seconds. No account, no card, no prior relationship needed. That's exactly why AI companies care. An autonomous agent can't open a bank account or pass a credit check, but it can sign a transaction. Google already built x402 into its own agent payment system. Cloudflare ships it by default in its agent toolkit. The actual usage is still small, about $24 million moved last month across 75 million payments, averaging 32 cents each. That's nothing next to what Visa or Mastercard move in a single day. But the average payment size is the tell. No card network on earth can process a 32 cent charge profitably. This protocol was built for a kind of commerce that doesn't fit inside the rails these same companies already own, which is exactly why they just joined it instead of competing with it.

• 
  Tai Groot 🐧 (@taigrr) reported 37 minutes ago

  So many people hate on @vercel / @Cloudflare for cost but they offer a really great service here People just really like building on nextjs. all in all it's good at being a framework Vercel and cloudflare both do a really good job with their wafs patching vulns before they are public and self hosters have a hard time keeping up, if they even know to.

• 
  Piyush (@PyDataWizard) reported 39 minutes ago

  Total monthly cost to run a startup: ~$20 - Claude ($20/mo) = coding - AWS Free Tier = hosting & DB - Vercel = frontend (free) - Stripe = payments (2.9%/txn) - GitHub, Clerk, Cloudflare, Sentry = all free Never been a better time to ship. What’s stopping you? 🚀

• 
  Adebayo (@debayoorr) reported 39 minutes ago

  Cloudflare is still down and somehow I've become a network engineer, a hostage negotiator, and a personal IT department, none of which I signed up for.

• 
  Chris baker (@derpy01) reported 39 minutes ago

  @marx1verstappen @SSolarite Sure, not with the concept of DNS, but any service that uses that DNS, which is cloudflare and runs most the internet, will not work properly

• 
  Rahul K (@rknkhanna) reported 42 minutes ago

  you mean 3 people are trying to cancel cloudflare?

• 
  Rob Terrin (@RobTerrin) reported 47 minutes ago

  @yrechtman Kernel access was really just useful for EDR (Crowdstrike and SentinelOne) and there have been multiple big exits since that gen like Cloudflare, Netskope, Rubrik, Wiz and Armis. If the labs lock down agent access, no cybersecurity company will become the next public company platform like Crowdstrike.

• 
  Sachin Joshi (@Sachin_is_here) reported 47 minutes ago

  Cloudflare also inserted itself at the perfect architectural layer: between users and origin servers. Once traffic already passed through its network, it could offer new products without asking customers to redesign their applications. CDN became the entry point.

• 
  Carqui 🐐 (@carquinyolis) reported 52 minutes ago

  @thecybersecguru @elhackernet In Spain whole cloudflare proxy IPs get banned with thousand of legitimate and corporate websites because just one seems to stream illegal sports. This is normal for non-technical government judicial powers..... Bad management

• 
  openwa (@openwadev) reported 53 minutes ago

  @mddanishyusuf @Cloudflare @Namecheap split between cf and porkbun. dont put all your domains in one basket, especially the one you use to log in to the domain service.

• 
  Cap-EO 👨🏾‍💻 (@EOEboh) reported 54 minutes ago

  The second real problem: backups With Postgres on a managed service, backups are a checkbox. But with SQLite, the database is just a file on my server, and if that server dies, so does my business 🥹 My fix: automatic snapshots shipped to Cloudflare R2 Cheap, but I had to build it myself (there are probably better approaches like a VPS backup).

• 
  Ernest Provo (@ernesttheaiguy) reported 57 minutes ago

  A 200 OK status does not mean complete data. Cloudflare found a race condition in hyper that truncated responses silently. Data leaders: never trust status alone. Verify payload integrity. #ResponsibleAI #DataStrategy

• 
  gökhan (@goekhan) reported 60 minutes ago

  fasted the whole day did some core work at home evening coffee gotta ship at least two other products today is "i just can't feel good if i want see that idea in action" for i am already paying for compute, cloudflare, hetzner, and thousand other APIs kinda funny though you can just function, worker and agent the entire internet into apps that talk to each other yet cannot watch a random world cup match from an ondemand commentator like Peter Drury or Alex Jacques and gotta help yourself with your mediocre homeboys you can have arxiv papers have their own podcast via notebookML but cannot push a likeness soccer commentator via kittenTTS or ElevenLabs API legally