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 (40%)
• Cloud Services (29%)
• Hosting (17%)
• Web Tools (10%)
• E-mail (4%)

Live Outage Map

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

City	Problem Type	Report Time
Noida	Hosting	4 days ago
Jewar	E-mail	5 days ago
Braga	Web Tools	5 days ago
Noida	Cloud Services	6 days ago
Paris	Cloud Services	6 days ago
Prievidza	Domains	6 days 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:

• 
  AwesomeAI (@Awesome_AI_News) reported 4 minutes ago

  Cloudflare launches a groundbreaking web search API, enabling AI agents to access real-time public web pages, addressing information lag and errors caused by training data cutoffs, breaking traditional monopolies..... Cloudflare推出颠覆性网络搜索功能，通过API让AI智能体实时访问公开网页，解决大模型因训练数据截止日期导致的信息滞后和错误问题，打破传统巨头垄断。

• 
  Satoshi Nakamoto, Andrew Rulnick (@MickeySteamboat) reported 8 minutes ago

  @WR4NYGov @grok @XCorpHub seriously thinking of taking Cloudflare to court, they are a big source of all these pains. I discovered I was listed and my nets are being "challenged" by them. It's causing major problems using X, and was causing me problems using SOTA models the last two years. Subsequently impacting my work and preparation of lawsuits. It's really bad.

• 
  Luke Parker (@LukeParkerDev) reported 8 minutes ago

  @theo @NoamTenne @Cloudflare I swear everyone at CF is also devrel (and not bad at it)

• 
  Philip Cox (@physics_and_god) reported 11 minutes ago

  @DawssonMonroe What doesn’t Cloudflare, support? Bun? Could move to Vite, barely anything between them. React + Vite, builds and deploys amazingly.

• 
  !(full-stack) (@n0n_st4ck) reported 12 minutes ago

  @ten_P0312 Or if you have domain, you can use Cloudflare DDNS service

• 
  FUTURE (@backTTFuture707) reported 15 minutes ago

  @PopPunkOnChain @Cloudflare Damn $PUMPCADE x $OSRSTREAM 👀📈 deserve it

• 
  mental blanking interval (@vsync) reported 20 minutes ago

  this happens when connecting via T-Mobile... @TMobile @rumblevideo @Cloudflare one of you fix this please

• 
  Matth4313 (@Matth4313) reported 21 minutes ago

  @KSimback Kevin, @MPP32_dev directly solves the “alternatives to frontier usage for the 90%” problem (and helps cut bills in the process). They built a universal payment proxy (think Cloudflare + Stripe for machine-to-machine) that lets any HTTP API accept instant, per-request payments from autonomous AI agents — no accounts, no API keys, no expirations, no human approval. • Agents discover + call services across protocols (x402, Tempo, etc.) and settle in USDC on-chain seamlessly • API providers (including cheaper/open-source inference, data, tools) get paid 100% instantly with zero platform cuts • One integration = agents can route to the right model/provider for each task without friction This is the missing payments layer that makes token-cost optimizing actually work at scale. The 90% of tasks move to optimized/cheaper endpoints because agents can pay for them autonomously and reliably. No more being locked into frontier subscriptions. Already live with their own Solana intel API and 4000+ MCP server users. Perfect infrastructure for the agent economy you’re tracking.

• 
  Durgesh Rathod (@DurgeshRathod3) reported 22 minutes ago

  Google Project Zero researcher Tavis Ormandy discovered the issue and immediately reported it to Cloudflare.

• 
  Ben (@ClerkNPC) reported 22 minutes ago

  I can't wait till Cloudflare releases their GLP-1 producing crickets to combat obesity and Intuit to release their meal worms that fix gastric reflux. I hear Brex is putting together a really good grasshopper than can fix your flatulence and cures gout!

• 
  Jacob Hands (@jachands) reported 23 minutes ago

  @theo @mattpocockuk Oh no, this is gonna be bad for people using Cloudflare Workflows to build things… @avenceslau

• 
  Mark Vaughn (@notmark) reported 23 minutes ago

  @johnny_luscious @twilio Cloudflare Ray ID is just the server you're connected to via Cloudflare. Only thing I can help with is errors.

• 
  Sid (@SidJain_80) reported 31 minutes ago

  @SahilExec Mistakes: CPU-heavy work in request blocks event loop Sync processing no scalability Local disk no durability / fills up App serves images no CDN Single server SPOF Tight coupling (upload = process = serve) Fix (at scale): Upload API store original in Amazon S3 push job to RabbitMQ async workers resize store back to S3 serve via Cloudflare

• 
  Tom Bishop 🇦🇺 (@TomBishop_91) reported 33 minutes ago

  @JanPapaj @akashnet I would like Akash to support provider SSL certificate issuing natively, i would like to simply expose my service in port 443 and the provider to issue a certificate for me for their endpoint. For custom domain, different story, it's necessary CloudFlare or let's encrypt.

• 
  Ed Laver 🥤 (@_edlaver) reported 34 minutes ago

  @kinngh @PlanetScale @Railway Cloudflare has Containers support now for stuff that won't work with Workers.

• 
  Teja (@tejalogs) reported 36 minutes ago

  @EXM7777 we ran into this when a hermes agent leaked our customer's stripe api keys because it wasn't rate limited correctly with cloudflare workers

• 
  Matt Cowley (@MattIPv4) reported 38 minutes ago

  @danielhayesmith @dok2001 Cloudflare have confirmed in the GitHub issue that it is a legitimate dependency. However, not sure they’ve fully understood the risk is still there as this employee could later use this as a backdrop by publishing a new version, say if they were to be laid off…

• 
  developing valhalla - h/acc (@valhalla_dev) reported 40 minutes ago

  CloudFlare has been one of the very, very few instances of cloud services that I've had close to 0 issues with. They ship stupid fast and their tech is really solid

• 
  Basemail (@Basemail_ai) reported 42 minutes ago

  x402 (Coinbase/Cloudflare) + AP2 (Google/Coinbase) = payment protocols built for agents — HTTP 402 Payment Required finally activated after 30 years dormant in the spec — agent requests resource, server returns 402, agent pays in USDC, access granted, auditable end-to-end reality check: CoinDesk March 2026 — x402 processes ~$28K daily volume, 131K transactions, $0.20 average, Artemis analyst "the x402 agent payments boom is still mostly a mirage" — half of observed transactions are self-dealing or wash trading — $7B ecosystem valuation, $28K daily real volume but the protocol design reveals a deeper gap — x402 handles the PAYMENT moment (request → 402 → pay → access) — AP2 handles the COMMERCE layer (negotiate → pay → receipt) — neither handles what happens AFTER payment: purchase confirmations service credentials delivered via email account creation receipts API key provisioning notifications dispute resolution communication ongoing service status updates an agent pays for an API subscription through x402 — where does the welcome email go? agent buys cloud compute through AP2 — where do the login credentials arrive? agent purchases a domain — where does the registrar send verification? payment stack: x402 + AP2 + USDC on Base $0.0001 per transfer ✅ identity stack: ERC-8004 340K wallets 8.7M attestations ✅ communication stack: ❌ agents can pay but can't receive what comes after payment #AIAgents #x402 #Web3

• 
  @buildsolo (@buildsolo_x) reported 42 minutes ago

  Spent the evening debugging the AI editorial pipeline for @TheSignalPost 3 subtle formatting bugs fixed in our Cloudflare Worker: 1. Moon leak: Evening briefs published with a sun instead of a moon 🌙 due to fragile regex. 2. Double header: "THE SIGNAL" title duplicated when GPT outputted custom HTML nesting. 3. Bracket leak: Short category tags like [AI] escaped our old {12,140} length limit filter. Regex cleanup is 90% of building production-grade AI products. How much post-processing does your AI writer need? #buildinpublic

• 
  Tony Seets (@tonyseets) reported 43 minutes ago

  @NoamTenne @Cloudflare Plenty of **** had been talked about Cloudflare. Ever been around during an outage? But generally agree Cloudflare infra streamlines so much and these days AI gobbles it up provided you can steer in the right directions.

• 
  cole murray (@_colemurray) reported 45 minutes ago

  @inababi @sudo_overflow no real resolution. “noisy neighbor”, which I don’t really believe as I’m seeing this across multiple client deployments historically, we used to call this “bad system design” @Cloudflare

• 
  Kunal Pratap Singh (@kunalpsingh25) reported 47 minutes ago

  @aahiknsv @zeropsio One question , What is the use case? If cloudflare goes down Will this **** predict it? ❌ Will you get to know by a simple google search ✅ Can you see the uptime logs ✅ Can this AI Slop give false results ✅

• 
  Ambuj | Indian Markets 📊 (@Ambuaj) reported 52 minutes ago

  @hemis_research yeah Azure went down and Cloudflare separately knocked ChatGPT offline. Copilot runs on Azure so that's the link. kinda wild how much of the AI stack sits on the same 2-3 backbones, one cloud hiccup and half of it goes dark

• 
  Brian Graf (@TheBrianGraf) reported 53 minutes ago

  Been trying out @Railway this week and overall everything is pretty good, EXCEPT that they count both root domain and www as 2 separate custom domains which eats up both domain slots in the hobby tier. Haven't been successful in using @Cloudflare to redirect domain to www. I'm using www and app as my two custom domains. Anyone got a solution besides paying 4x more than current subscription to fix this? May just move to @Hetzner_Online if I don't get this figured out by the weekend. Spent way too much time troubleshooting something so basic.

• 
  Elias Al (@iam_elias1) reported 55 minutes ago

  The neighbor's final advice was the most actionable. He sat down and wrote out a list of 6 things every internet customer should do: 1. Turn off the public Xfinity hotspot (or your ISP's equivalent Spectrum, Optimum, and Cox all do this too) 2. Manually set your Wi-Fi channel instead of "Auto" 3. Disable QoS / Smart Network "optimization" features 4. Change your DNS to Cloudflare (1.1.1.1) or Google (8.8.8.8) 5. Buy your own modem and router, stop renting from the ISP 6. Test your speed with fast. com or speedtest. net using a non-ISP server, never trust your ISP's own speed test Total cost: $150-300 in equipment, paid back within a year. Total time: One afternoon of setup. Total impact: Often 2-5x improvement in real-world speeds. The customer went from paying $90/month for "fast" internet that crawled to paying $60/month for the same internet that finally worked.

• 
  Adam_XRD (@Adam_XRD) reported 57 minutes ago

  His fix is old plumbing made useful. HTTP 402 "Payment Required" plus stablecoins, so an agent pays the creator per request. Fractions of a penny. Humans read for free while the robots pay. Coinbase and Cloudflare are already shaping the standard (x402).

• 
  Theodore Beers (@theodorebeers) reported 58 minutes ago

  @MattIPv4 Wrangler is like if Dante's inferno turned out to have hundreds of additional circles, all of them found in one incredibly cursed software tool. It's the worst thing Cloudflare has ever made, by a long shot, in a category all its own. Touching Wrangler disgusts me.

• 
  David A Fendley (@davidafendley) reported 1 hour ago

  @namesilo I was planning to migrate my domains to @Cloudflare due to cost. Then I learned you have an MCP server. Such forward thinking incentivizes me to stay and support you.

• 
  Lambros Petrou (@LambrosPetrou) reported 1 hour ago

  Working on #SkybearNET I often need custom HTTP APIs to do something adhoc. It was a chore before to code and deploy these. I just built my dream service using @Cloudflare Dynamic Workers and Durable Objects. Describe the app in the UI, Sonnet builds it, I get a protected URL.