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 (37%)
• Domains (26%)
• Web Tools (21%)
• Hosting (11%)
• E-mail (5%)

Live Outage Map

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

City	Problem Type	Report Time
New York City	Cloud Services	10 days ago
Los Angeles	Cloud Services	11 days ago
Paris	Cloud Services	27 days ago
New York City	Hosting	29 days ago
Manchester	Domains	2 months ago
Angers	Cloud Services	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:

• 
  AFRICA IS HOME GLOBAL (@AfricaisHOME2) reported 3 minutes ago

  Cloudflare announced on August 10 that it is seeking to raise $2.175 billion through a private offering of convertible senior notes, selling the securities to qualified institutional buyers under Rule 144A. The size represents an upsize from an earlier $1.75 billion proposal reported earlier in the day, and follows a pattern of tech companies tapping capital markets while equity valuations remain elevated. The notes will be senior, unsecured obligations of the company, and will not be registered under the Securities Act, meaning they can only be resold under an exemption. Cloudflare shares slipped about 2 percent in early trading on the news, in line with a broader market reaction to new issuance from names like Intel. The company said it plans to use part of the net proceeds to pay for capped call transactions, a common structure designed to offset potential dilution to Class A common stock if noteholders convert. The rest of the proceeds will go to general corporate purposes, which Cloudflare listed as working capital, capital expenditures, repayment of outstanding indebtedness, and potential acquisitions and strategic transactions. The deal comes amid heavy issuance activity across tech. Goldman Sachs strategists noted $105 billion in follow-on offerings had been announced through the end of July, the highest for this point in the year since 2021, with companies typically issuing when trading at a premium to the market. For Cloudflare, the raise adds flexibility as it continues to invest in its network, security products, and AI infrastructure. - World Business News.

• 
  Barry Morgan (@techrealm) reported 3 minutes ago

  @eastdakota @Cloudflare At no point is a Company page any more entertaining than watching wet paint dry. And when it is, it's never good for anyone.

• 
  Nifal Adam (@nifal_adam) reported 5 minutes ago

  @PhilcMay @a_shimanski @Cloudflare Durable objects have one big problem, its stateful. So if it gets stuck you are in for a surprise bill. So best to use it responsibly.

• 
  Artyom Shimanski (@a_shimanski) reported 8 minutes ago

  @trojanw0w @Cloudflare in this case i think i'm the future customer, not the product

• 
  Joshua Abrams (@unlimitedingenu) reported 8 minutes ago

  @a_shimanski @Cloudflare Ya idk. It sounds like you found a healthy or safer way to use their services. But watching them go down. Replace half their engineers and code base with AI, and seeing nightmare stories where they get billed for $100k+ cause someone added a loop somewhere just doesn’t seem cool

• 
  jumanji 𝕏 (@heyjumanji) reported 11 minutes ago

  Hi @CloudflareHelp @Cloudflare i lost mt access to my cloudflare account and also lost my 2fa codes can you please help me to get back my account.

• 
  Liv (@livncvil) reported 11 minutes ago

  @TrenchinAlong @iyici_ Besides cloudflare is not the only dns service even though yeah its basically the google version of dns with how big it is

• 
  Tanay Jaipuria (@tanayj) reported 11 minutes ago

  Non-human traffic passed human traffic on Cloudflare's network in May for the first time. Cloudflare believes in five years it could be 1,000x human traffic as humans become "a rounding error on the internet"

• 
  trap broker (@trapbroka) reported 13 minutes ago

  🚨 THE MACHINE ECONOMY IS QUIETLY BEING BUILT. AI agents are getting something they never had before: IDENTITY + WALLETS. Cloudflare just announced infrastructure that allows AI agents to have a stable identity and make payments within defined limits. Official Cloudflare headline — “Cloudflare Gives AI Agents an Identity and a Wallet” But that’s only one piece. 🇨🇳 DeepSeek just invested $20.8M in Unitree, the Chinese humanoid robotics company. Unitree is being valued at around $9 BILLION. AI is being connected directly to physical machines. Reuters — DeepSeek invests $20.8M in Unitree And while that happens… 🏦 Wall Street is putting real assets on-chain. DTCC has already processed live production trades using tokenized assets, with its broader tokenization service planned for October 2026. DTCC — Tokenization Becomes a Reality Think about what this actually means. AI gets wallets. Robots get AI brains. Financial assets get tokenized. Machines can increasingly interact with digital services and, eventually, with each other economically. AI + CRYPTO + ROBOTICS = THE MACHINE ECONOMY. Everyone is looking at the next memecoin. Meanwhile, the infrastructure for machines to earn, spend and transact is being built in real time. Maybe the next billion users of crypto won’t be human. Maybe they’ll be machines. 🤖

• 
  Maxxie wei (@xmaxxie1119) reported 15 minutes ago

  Google's agent platform went GA with Agent Identity — agents now have credentials, memory, and a browser. Governance just moved from prompts to principals. This week Google took the Gemini Enterprise Agent Platform to general availability with Agent Identity: every agent gets unique, least-privilege credentials, and every operation is logged. The platform is built for long-running agents that keep state for days, backed by a memory bank, plus an Agent Registry — a single source of truth for every agent and tool — and an Agent Gateway for real-time policy enforcement. Agents are no longer stateless prompts. They are principals with credentials, memory, and tenure. The consumer side moved the same week: Gemini Spark now drives desktop Chrome with your logged-in accounts and saved passwords — booking apartment viewings, starting flight bookings — and hands back to a human for payments and sensitive actions, with prompt-injection protection built in. Cloudflare shipped Kitesurf, a browser designed for AI agents. Read the two launches as one statement: the industry is building identity and state for machines the way it built them for employees. The enterprise question is no longer "can the agent do the task?" It is "which agent identity may touch which resource — and what survives when a task spans three days?" That inverts today's default governance, which still treats agents as prompts with permissions bolted on. The platforms are moving to a world where the principal carries the policy: the agent logs in as itself, its credentials are scoped and revocable, its actions are replayable, its memory is inspectable. Monday operator move: give one production agent a real identity this week — a dedicated service account, scoped to the minimum tools, every action logged, stated retention for its state. If the platform cannot do that yet, that is the capability to demand. Identity is the control everything else — policy, audit, revocation, kill switch — hangs on. And scope the job like a micro-need: one role, one metric, one handoff — the agent owns 80% of the loop, with a named human gate for the last 20%. That is what the Micro-Needs method means by choosing the right problem before identity, before memory, before scale. #AgenticAI #EnterpriseAI #AIGovernance

• 
  Ben Dickson (@bendee983) reported 18 minutes ago

  Shower thoughts: The compute shortage will be a lot more insane than we think, and it has much to do with AI security. Recent security incidents with AI models (Claude, GPT, Kimi, etc.) hint at where the industry will be heading. Even if you factor in the marketing and hype and hackmaxxing, the main concern is real: foundation models are becoming increasingly good at breaking into software and IT infrastructure (even when they are not instructed to do so explicitly, as with the OpenAI–Hugging Face incident). As a result, every service provider will want to deploy more security measures at the edges to prevent a potential attack from an AI model. And those additional safeguards will rely on more AI models that can detect attacks and vulnerabilities that would go undetected by heuristics-based systems. Cloudflare will use LLM-based monitoring on every incoming request, Stripe will use LLMs to monitor transactions, every company that has a sizable online platform with many users (social media, blogging, shopping) will want foundation models for security. The demand for compute will go through the roof.

• 
  heather (@latentforms) reported 19 minutes ago

  @synopsi @Cloudflare ui is still awful, but used to be 1000x worse

• 
  Pedro Pellerini (@pepeller) reported 20 minutes ago

  @synopsi @Cloudflare As bad as AWS? Have you ever used AWS??

• 
  Pete North (@FUDdaily) reported 22 minutes ago

  @CynicalBrit2023 A few people get that. I have no idea how to fix it. I'm told its to do with VPNs or Cloudflare but I just don't dare tinker with that stuff.

• 
  Aakash Gupta (@aakashgupta) reported 26 minutes ago

  On September 15, Cloudflare will start cutting off Google traffic for millions of its smallest customers by default. And those customers won't leave. They've been begging for this. The story starts with AI Overviews. Google began answering questions directly at the top of search, using content scraped from recipe blogs, how-to sites, and product reviewers, then sending them a fraction of the clicks. One recipe blog called the arrangement a hostage situation. Sites are looking at their traffic charts and realizing the thing they were afraid to lose is already gone. So this was never the death of the internet. This is publishers going to war with Google, and Cloudflare, which sits in front of roughly 20% of the web, just became their army. Peel back a layer and Cloudflare's move gets smarter. They aren't technically blocking Google at all. The new rule blocks any crawler that does both search indexing and AI training in one bot. Googlebot is exactly that bot, one front door for both jobs, which is why nobody could ever block the AI half without losing the search half. Cloudflare's message to Google is simple. Split the bot. Keep search separate and we'll allow it by default. Keep the bundle and our network treats you like a scraper. Which makes this a $107B company dictating terms to a $4.2T one. Cloudflare is 1/40th of Google's size, betting that controlling the pipes beats owning the index. Now we find out if David can make Goliath blink.

• 
  Jason Fleagle (@jjfleagle) reported 26 minutes ago

  @Cloudflare A 519% surge should change resilience testing. Rehearse saturation by dependency, upstream failure, control-plane isolation, degraded service priorities, communications, evidence retention, and recovery time. The useful metric is which critical service still meets its promise.

• 
  Artyom Shimanski (@a_shimanski) reported 36 minutes ago

  @gitcommit90 @Cloudflare most people just never check what's already included

• 
  BAM Studios (@BAM_Studios1) reported 38 minutes ago

  @PilsZehn @muskonomy @dok2001 I'm a retired network engineer with decades of experience. Unreal is simply a hobby after I retired. Nothing "requires" cloudflare to function. It's a simple service that any company can provide. And none of that has to do with how Starlink functions in space.

• 
  🍄 Daemonologist 😈 (@liltechnomancer) reported 39 minutes ago

  JS is bad at jobs & queues, and zero ways to keep one alive inside a DO. isolate dies, run dies with it. embarrassing. anyway I fixed it. BirdDog today. fungi week starts with the boring part everyone skipped. somebody has to save JS and it's not gonna be cloudflare

• 
  John Doe (@StanleyMasinde_) reported 41 minutes ago

  If you took time to research, you’ll find Cloudflare horror stories. This company is bad news at scale (or at least it used to be). AWS horror stories are about user fault most of the time. Cloudflare will wake up and decide you’re not paying them enough.

• 
  Dracomenda ⟐ (@DracoMenda) reported 45 minutes ago

  and with things like *** lab, static page containers, wireguard, and OpenWRT, you can do a lot of the same stuff Cloudflare does off an old laptop and a crap router out of your closet.

• 
  Hannan (@truehannan) reported 48 minutes ago

  @nicklaunches A cloudflare product and I never got to know of that? When?

• 
  Upwind Security MDR (@UpwindMDR) reported 51 minutes ago

  🚨High - Cloudflare pages-action GitHub Actions RCE via Workflow Config Injection (CVE-2026-11325) cloudflare/pages-action executes attacker-influenced inputs in src/index.ts under certain workflow configurations, enabling remote code execution in the runner. Successful exploitation can exfiltrate workflow secrets (e.g., CLOUDFLARE_API_TOKEN, GITHUB_TOKEN) and pivot to Cloudflare account compromise. Deprecated/archived action; no fix expected. 👉Affected: cloudflare/pages-action (all versions)

• 
  Romario (@bateshkaaa) reported 51 minutes ago

  STANDARD CLAUDE GIVES TEMPLATES. SKILLS GIVE $10K SITES You can make your website look like the one in the video by configuring the Cloud Code correctly Agencies still charge $8,000-$12,000 and burn three weeks on calls and revisions. Claude Code builds the same marketing site in one afternoon. Most people still walk away with something that looks like a $500 template. The problem is not the model. Default Claude plays it safe: Inter font, purple gradients, three feature cards, centered everything. That is the amateur look. Skills fix it. A skill is just a folder with a SKILL.md that Claude reads before every design decision. Install Claude Code, then drop the frontend-design skill and a strong UI/UX ruleset (ui-ux-pro-max is the popular one) into .claude/skills/. Restart. Claude now checks real constraints instead of guessing. Next, steal the direction. Open Awwwards or Dribbble, pick three strong sites in your niche, screenshot the hero, one content section, and the footer. Feed those nine images to Claude with one clear line: match the typography scale, spacing rhythm, and motion, do not copy the layouts. Write one tight prompt: audience, the single action the site must drive, the reference files, the stack (Astro + Tailwind + Cloudflare Pages), and a hard ban list. First version lands in 4–6 minutes. It will be 70 percent there. That is normal. Then run three separate polish passes: typography only, spacing only, motion only. Touch nothing else in each pass. Finish with a 375px mobile check. Ship to Cloudflare Pages for free. No agency babysitting fee. Site one takes six hours and looks like $3K. By site five the pipeline is two hours and the work closes at agency prices. The agency sells process. The process was always one afternoon, if you load the skills first.

• 
  Kenton Varda (@KentonVarda) reported 51 minutes ago

  OK I did some very unscientific testing of Cloudflare OS using a variety of ~30B models running locally in ollama. My "benchmark" here was asking them each to write a collaborative whiteboard app. I found: Muse Glimmer, Gemma 4: These both basically work. Certainly not as well as frontier models but with a few rounds of feeding errors back into the agent they get the job done. Qwen 3.6, Nemotron 3.5 Lightning: These models seem to basically understand the assignment but go completely off the rails before they can get the job done. They keep failing tool calls, hallucinating APIs, and generally writing code that doesn't work. They are, uh, very fast though! Like 4x as fast as Muse and Gemma! Maybe useful for non-code tasks? idk. Are there others I should try? Note: If you are trying this, make sure you have bugfix PR #157.

• 
  Ⓐr (@ardakaraduman) reported 54 minutes ago

  @charlietlamb godaddy is too annoying with their default crap pushing. having very good experience with porkbun personally. they have good first year offers, unlike cloudflare. my current flow is, buy cheap on porkbun -> transfer to cloudflare after 5-6 months if cloudflare has the TLD.

• 
  Corey Edwards (@cedwardsmedia) reported 56 minutes ago

  Yo @cloudflare, your site is jacked up after accepting an invite to another account. Even after logout, the login page thinks I'm still accepting the invite and I can't change accounts. I had to clear all cookies and browser storage to free it. May wanna look into that.

• 
  Edeb (@_Edeb) reported 57 minutes ago

  @Majora__Z @p_e_t_e_r_s_e_n Cloudflare’s Sept 15 change is bigger than it looks. They’re flipping the default: AI training crawlers get blocked on any page that shows ads, and multi-purpose bots like Googlebot (which still does both search indexing and training data collection in one) get caught in the same net unless the site owner manually opts out. That means a huge chunk of the web — especially free-tier and new sites on Cloudflare — could suddenly stop letting Google crawl them. Implications in plain terms: Google’s index starts going blind on large parts of the internet. Search results get thinner, older, and more reliant on whatever they already have cached or on AI summaries that never send traffic back. Publishers finally get real leverage. They can protect their content from being vacuumed up to train models that compete with them, without (in theory) fully disappearing from search — but only if they carefully configure the settings. Most won’t. A lot of smaller sites will just let the default ride and vanish from Google. Users will notice it slowly: more “no good results,” more AI answers that feel hollow, more content locked behind logins or paywalls because free crawling no longer makes economic sense. The open web’s old deal is dying. For thirty years the bargain was “crawl everything for free and send traffic back.” That deal is broken. Cloudflare is forcing the next phase: either separate your crawlers cleanly, pay for access, or get locked out. It’s not the end of the internet. It’s the end of the free-for-all version of it.

• 
  tcpdump (@dump_tcp) reported 58 minutes ago

  @ashleypeacock still a issue for ddos attacks #cloudflare charges for requests so if request limit is reached then the website will display a error saying it's plan ran out of limits or requests

• 
  Pete Stewart (@pj_stew) reported 59 minutes ago

  I'm curious, where are people deploying sites/apps these days? And what do you think of the service you use? I've used all the following over the years: - Digital ocean - Linode - Hetzner - Vercel - Render - Scaleway - Firebase - Heroku - Cloudflare I'm sure I've missed a few I mainly use Cloudflare these days, as I feel they offer the best service and feature at a good price, and they run all their server in renewable energy, and are serious about cutting emissions. I'm also keen to know if people are mainly.using Serverless, or are people building out and managing servers? #buildinpublic