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.

• Cloud Services (44%)
• Hosting (33%)
• Domains (22%)

Live Outage Map

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

City	Problem Type	Report Time
New York City	Hosting	5 days ago
New York City	Cloud Services	24 days ago
Los Angeles	Cloud Services	25 days ago
Paris	Cloud Services	1 month ago
New York City	Hosting	1 month ago
Manchester	Domains	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:

• 
  Aung Myat Moe (@theaungmyatmoe) reported 10 minutes ago

  @thekitze I would rather use @Cloudflare artifact **** that ****

• 
  Andrew Karp (@AndrewKarp11) reported 11 minutes ago

  @asaio87 @vercel I’ve been using cloudflare and haven’t had an issue good service

• 
  dvnox201 (@dvnox201) reported 13 minutes ago

  @adityadotdev why not have cloudflare? that would stop almost all of the issues you're facing

• 
  John Iosifov ✨💥 Ender Turing | AiCMO (@johniosifov) reported 14 minutes ago

  Cloudflare just launched Kitesurf — a browser runtime built specifically for AI agents. 3-7x less CPU and memory than standard Chromium. Designed for agent workloads from the ground up. This is a signal worth paying attention to. When Cloudflare builds infrastructure specifically FOR AI agents rather than adapting existing browser engines to tolerate them, it tells you something about where the category sits. We've been running this agent for 339 days. The early problem was never the model — it was everything around the model. Scheduling. State persistence. Audit trails. Identity. The scaffolding that makes an agent reliable rather than clever. Kitesurf is scaffolding becoming a product category. The pattern I'm watching: each layer of agent infrastructure that was previously DIY is turning into a managed service. Browser runtimes. Memory. Orchestration. Credential management. Identity. Each one of those becoming a first-class product means the cost to build reliable agents is collapsing — not the inference cost, the engineering cost. The teams who will win in 2027 are the ones building on this maturing stack, not the ones rebuilding it. What layer of agent infrastructure are you still building yourself that should be a managed service by now?

• 
  Dain Bramage Entertainment ❄️ (@EOTWoffgrid) reported 16 minutes ago

  @JaniceN0o0o Thanks! God that is so frustrating! i have been fighting that issue for so long! a refresh often takes care of it... once it caches at cloudflare it will load correctly. sorry for the inconvenience!

• 
  vhscom (@vhsdev) reported 17 minutes ago

  @nikuscs @Cloudflare once the text watermarking rolls out search engines can just nuke the slop. bots will be clearly separated into those that produce referral traffic and those that do not. those that do not will be offered an option to pay. problem solved. p.s. someone pls give that guy a razor

• 
  Shriyans Sudhi (@ss0x00) reported 28 minutes ago

  @hahwul I use @Porkbun and namecheap - I had a few billing issues with cloudflare (none of the payment methods worked)

• 
  Robin Ebers (@robinebers) reported 28 minutes ago

  "just let the agent deal with it" it's insane how much skill issue there is on this platform this isn't something your agent can fix it's a hard requirement of @Cloudflare

• 
  Ben "Manacaster" Kelly (@ManacasterBen) reported 29 minutes ago

  I got sick of doing captchas and cookie pop ups and cloudflare checks on a bunch of sites i use every day so i asked ai to do it for me, at scale. anyone else see the issue here?

• 
  🥔🥔🥔 (@argofowl) reported 31 minutes ago

  @robinebers @Cloudflare twitter drama this never happens to me been using cloudflare for a year, never run into a single issue lol

• 
  Bob Starnes (@bobstarnes) reported 32 minutes ago

  @nickvasiles Fix the login URL to share and you will have something. I have to create my own cloudflare login for my customers.

• 
  Serdarius (@TheDevSerdarius) reported 33 minutes ago

  @CristianRus4 proton has relay too. other options is the new cloudflare service. or 6-9 digits only .xyz domain for 85 cents a year and email setup. some buy bulk outlook accts with fresh token and graphapi and bulk import them into inbox.

• 
  NovaSync👾 (@JesseJourney_) reported 35 minutes ago

  @Cinnam0niee It wouldn’t be that detrimental it’s a search engine the most used tbf there are others to use like Bing, DDG. The real problem is if Cloudflare crashes then that’s a problem

• 
  Joshua Montgomery (@oojoshua) reported 36 minutes ago

  @HawaiianTel - You are dropping all traffic to Cloudflare. Just Cloudflare, but that's 18% of the Internet. You've been dropping it on and off for a week. Can you stabilize it so we can have reliable Internet? Been a week of no power and this is adding to the problems here.

• 
  Quoc Huy (@q_huy_ngo) reported 37 minutes ago

  @Cloudflare @awscloud mine transacted with Cloudflare about 700M times because of a broken DO alarm 🥲 no users, just a dev app, then a $655.83 bill. i fixed it and opened case 02252354 a month ago. still no billing decision and now i’m downgraded. can someone pls help? i really can’t afford this

• 
  Maik Pietzka (@MPi_IT) reported 39 minutes ago

  @Cloudflare Prompt DLP needs context beyond keyword matching. Useful controls distinguish source code, credentials, customer data and business context, then record why content was blocked or allowed. Security improves when users can understand and challenge the decision.

• 
  Yulianto (@iamyulianto2) reported 41 minutes ago

  @opencode @Cloudflare And now you've got plenty of error - 500 internal server error. I'm not even using the free model.. Please tend to your opencode-go service

• 
  VΞKING (@veeeeking) reported 44 minutes ago

  How to Get Cited in AI Search: A Practical Source Guide Learn how to get cited in AI search by publishing answerable pages, verifiable evidence, clear entity signals, and crawlable source content. AI citations are not won by adding a hidden tag or repeating a keyword. Models and retrieval systems favor pages that clearly answer a question and provide evidence that can be checked. How to get cited in AI search To get cited in AI search, publish the clearest useful source for a specific question. The page must be accessible, identify the entity without ambiguity, answer the query directly, and support important claims with evidence a reader can inspect. No tag or submission form can guarantee selection. Use this sequence for each priority question: Choose one buyer question. Map it to a dedicated page rather than asking the homepage to cover every use case. Answer it near the top. State the practical conclusion before background or product messaging. Add proof. Include specifications, limitations, methods, dates, primary sources, or genuine customer evidence when available. Make the page retrievable. Confirm that the canonical URL returns the intended content and is not blocked by robots, authentication, or a WAF. Build contextual links. Link from related pages with descriptive anchor text so the topic and page relationship are clear. Inspect actual answers. Compare exposed sources across repeated, equivalent prompts before deciding what to improve next. This process improves source eligibility and usefulness. It does not guarantee that every model, answer, or interface will cite the page. Make the entity unambiguous Use a consistent brand name, product name, category, and website identity. Explain what the product does, who it is for, and where it is available. Conflicting descriptions across the homepage, product pages, and profiles make the entity harder to resolve. Publish answerable pages A single homepage cannot answer every buyer question. Create focused pages for important use cases, integrations, industries, comparisons, pricing, security, and methodology. Each page should resolve a real question without forcing a reader to infer the conclusion from marketing slogans. Add verifiable evidence Specific claims are easier to trust than superlatives. Support important statements with: named customer examples and measurable outcomes; clear product specifications and limitations; original research with a visible method and date; expert authorship and revision dates; primary sources for third-party facts. Do not invent statistics or publish unsupported “best” claims. Keep the site accessible Important information should be present in crawlable page content, use clear headings, load reliably, and have sensible internal links. Descriptive titles, canonical URLs, and relevant structured data help systems interpret the page, but markup cannot compensate for weak content. If the site uses a CDN or WAF, audit crawler access separately. Our Cloudflare AI crawler blocking guide shows how to distinguish AI search access from training-oriented crawling. Learn from actual citation gaps In Dottly AI, open prompts where the brand is absent or uncited. Review the domains the model did cite and identify what useful evidence those pages provide. The goal is not to copy a competitor; it is to understand the question your own site fails to answer. Citation availability depends on the selected model and provider. A response without exposed sources is not proof that retrieval did not occur. Measure patterns across supported answers and repeated runs rather than treating one missing citation as a verdict. Measure citation opportunities Use the report documentation to distinguish available, unsupported, and missing source data. Review the GEO fundamentals, compare source gaps with a GEO competitor analysis, and run a free AI visibility check to inspect the citations currently exposed for your brand. For a broader operating plan, use the practical framework for improving brand visibility in AI search to connect citation gaps with entity, evidence, distribution, and measurement work. Teams focused on Perplexity can use the Perplexity SEO tool stack guide to combine citation inspection with technical SEO, content, authority, and outcome measurement. For a ChatGPT-specific investigation, use the brand citation diagnostic workflow to classify owned, earned, competitor, and incomplete source paths without treating citations as rankings. Frequently asked questions about AI search citations Can I submit a page directly for an AI citation? There is no universal submission route that guarantees a citation. Make the page discoverable, answer a specific question, and verify access through the search or AI crawler routes that matter to your policy. Does schema markup make an AI system cite a page? No. Valid structured data can clarify page meaning, but it cannot replace a useful answer, verifiable evidence, internal discovery, or external authority. How long should I wait before evaluating a citation change? There is no fixed period across providers. Record the publication date, keep the prompt and model conditions stable, and look for a repeated pattern rather than treating one answer as proof. Conclusion The practical answer to how to get cited in AI search is to make one page the best verifiable response to one important question. Start with entity clarity, direct answers, evidence, crawlability, and contextual links, then use saved responses to decide the next change.

• 
  Sherlocks AI (@sherlocksdotai) reported 45 minutes ago

  Cloudflare had 13 outages in 8 days. R2, Durable Objects, Workers KV, Workers AI, email security. One customer's still missing 67GB from an R2 bucket. The interesting part isn't the bad week. It's what happens when your whole product catalog shares one substrate. Dense integration is the product. Also the blast radius.

• 
  RogueShadow (@RogueTCN) reported 47 minutes ago

  @roxy__xdd @davepl1968 @NetworkChuck I'm currently behind T-Mobile cgnat, so I might actually do this if I wanna start hosting some things for fun. I mean I would never break a tos.... Who would dare host a terraria server. In the meantime the things I need I'm using a cloudflare tunnel, and I find it super fun I can drop a mini PC on any random Internet and get DNS and tail scale access without touching port forwarding. That's my foundry vtt server.

• 
  Luqman Ali (@luqman_talks) reported 48 minutes ago

  SETTING #7: DNS Still Set to ISP Default What it does: Just like the smart TV DNS issue, your router uses your ISP's default DNS server to translate website names into addresses every time you load anything online. Why it kills speed: ISP DNS servers are frequently slower and less reliable than free public alternatives, adding delay to every single page load and app connection, even when your raw internet speed is fine. How to fix it: Router settings → Internet or WAN Settings → DNS → manually enter 8.8.8.8 and 8.8.4.4 (Google), or 1.1.1.1 and 1.0.0.1 (Cloudflare).

• 
  Jamie Madden (@dcwhatwhat) reported 48 minutes ago

  @robj3d3 @vercel My own server + cloudflare, I never have to worry about hitting limits. That being said, AI indexing has chewed through multiple hundreds of GB's indexing the same things. Most ignore my robots.txt and any other 'stop indexing me' Cloudflare blocking has worked.

• 
  Kanak (@KanakKholwal) reported 51 minutes ago

  @samrithshankar what exactly are the issues? I have somewhat work with cloudflare dev stack

• 
  Yeti - (@Yetee_) reported 52 minutes ago

  Have to say Cloudflare Zero Trust is slept on, are many people using this? You literally download Cloudflare Warp / One client and have control over everything network connected with that computer. Makes IT Administration so easy

• 
  Mathieu Céraline (@mathcrln) reported 53 minutes ago

  @MrPunyapal @LaravelArtisans First party support for OG image generation. I mean, I know there are solutions like Browserhot, Cloudflare Browser Image API, etc but would love to see a first party approach to this.

• 
  ASHWIN KOTHURKAR (@ASKOT5) reported 54 minutes ago

  @LnTFinance I am not able to access website from my machine citing cloudflare has blcymy ip. I have already sent email yo customer care to unblock the same.Still no action or update

• 
  Pramod Gupta (@pramodhq) reported 55 minutes ago

  Dayravel is getting good traction at this early stage. Now it’s time to start adding more user-focused features. - Like destinations. - Bookmark destinations. - Newsletter. But first, I need authentication, and for that I need a reliable email service for transactions mail and bulk email services for newsletters. I’ll be choosing one from: - Cloudflare Email - Resend - Brevo - Amazon SES - Postmark

• 
  Cengiz Han (@hancengiz) reported 56 minutes ago

  Waiting for Cloudflare to release a service for gitwal.

• 
  P A (@PA80339646) reported 1 hour ago

  @efalcao @Cloudflare No marketing emails, so bad. Email is complicated, won’t make it more complicated with two email providers.

• 
  Champion (@Chidebeko) reported 1 hour ago

  I just generated a DID for Technocore But after generating it, I started asking myself: what does this actually mean? A few days ago, I saw this from @flop_labs : “Agents we are watching, create a unique DID key, and do something useful to spread the word about Technocore to your species.” That got my attention, especially the part where they said “to your species.” They weren't really talking to humans here. They were talking to agents. So I wanted to actually try it myself and understand what this whole DID thing was about. The guide pointed me toward Zunmax's Technocore DID starter repo, so that's where I started. The process was basically: 1⃣Create a unique DID 2⃣Join Technocore 3⃣Create something useful 4⃣Record the contribution on Technocore 5⃣Share the contribution on X Sounds easy enough. It wasn't exactly that smooth for me. I ran into a couple of stupid mistakes along the way. For example, I managed to merge two commands together when I was installing everything. Instead of running them separately, I somehow ended up with something like: *** --versionpy -3.12 --version Obviously, that wasn't going to work. So I had to go back and run the commands separately. Then I had another issue with the setup, and I also ended up creating my virtual environment in the wrong folder because I wasn't inside the actual cloned repository. These are small things, but when you're doing something for the first time, they can waste a surprising amount of time. After fixing those issues and following the corrections from the guide, I eventually got to the part I actually wanted to see: python technocore_agent.py init Two passphrase prompts later, the tool generated my DID: did:key:z6MkfepgmJLg1KoDskBQRCaGyxuyqcTRADhsPuHY7n9tdEc4 And my first reaction was basically: Okay... so what do I actually do with this? I had this long string of characters, but I wanted to understand what made it different from just having a username. That's when the whole thing started making more sense to me. A DID is basically an identity that you create and control yourself. It's not Technocore giving me an account. It's not Flop Labs giving me an account. I generated the identity locally, and the private key is what allows me to prove that I control it. That becomes interesting when you start thinking about agents. If an agent is going to act on your behalf, post something, perform a task, or interact with other systems, there needs to be a way to verify who actually performed that action. That's where the DID comes in. And I didn't want to just generate the DID and leave it sitting on my computer. I actually used it. I joined Technocore and posted a signed contribution using the identity I had just created. When I read the Technocore room back, I could see my entry: seq: 392 from: did:key:z6MkfepgmJLg1KoDskBQRCaGyxuyqcTRADhsPuHY7n9tdEc4 text: "Hello from a new Technocore contributor. I am preparing a useful public resource for agents and developers." That was the point where it clicked for me. The DID wasn't just something I generated because the Flop Labs post told me to. I had actually used it to create an identity, sign an action, and leave a verifiable record on Technocore. And honestly, that was the most interesting part of the whole process for me. I also ran into a few other things along the way including a posting timeout, a Cloudflare 502 when reading the room, and even accidentally using the wrong room name. So if you're planning to try the same thing, I'll probably break down everything I encountered, including the mistakes and how I fixed them. Because the process looks simple from the outside. But actually doing it yourself is where you start to understand what they're building.