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 (39%)
• Domains (22%)
• Web Tools (22%)
• Hosting (11%)
• E-mail (6%)

Live Outage Map

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

City	Problem Type	Report Time
New York City	Cloud Services	11 days ago
Los Angeles	Cloud Services	12 days ago
Paris	Cloud Services	28 days ago
New York City	Hosting	1 month 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:

• 
  Peter Machona (@chilumbamachona) reported 9 minutes ago

  @AnthropicAI @OpenAI I.e you ask Claude to help you decide between Cloudflare AI gateway & OpenRouter - it will search the web and come back with a comparison table that actually has some values that are false and until you check you would not know.

• 
  Joey Chilson (@joeychilson) reported 9 minutes ago

  Railway is the agentic cloud, not Vercel or Cloudflare. I can prompt an agent from my phone to build anything I want and have it deploy it and give me back a domain. No restrictions. If anything happens the Railway mobile app tells me and I just prompt the agent to fix.

• 
  Hamid Siddiqui (@hamidInventions) reported 10 minutes ago

  @wickedguro @markgalkev tbh now not 100% sure if it might be the same reason since you mentioned everyone getting this same time but i strongly feel it would be Cloudflare only, I faced this issue few months back. When i was debugging this I disabled Cloudflare proxy for a while and tested posting the posts that were failing to post on TikTok, and they worked fine. I also used to face similar issue with fal.i/replicate where they would randomly fail to get the R2 image url i passed for image generation. It was fixed as well. The exact steps i took I don't recall, would have to check what settings i did change eventually.

• 
  stackzz (@stackzz) reported 10 minutes ago

  🌐 Cloudflare’s Two Filters Broke Delivery Cloudflare Email Security posted a major incident at 8:04 AM PDT: customers using Spamhaus filtering downstream were rejecting mail from 134.195.26.0/23. Cloudflare says those IPs should be allowed and recommends disabling that downstream Spamhaus filter while it investigates. That’s a brutal operator lesson. Two security layers can each look reasonable, then collide and turn valid mail into a false positive. If trade alerts, password resets, client approvals, or incident pages depend on email, “sent” is not delivery proof. Trace the whole handoff: source accepted, relay accepted, recipient accepted. Anything less is a green dashboard hiding a broken path.

• 
  Ashley Peacock (@ashleypeacock) reported 13 minutes ago

  I don’t use a VPS at all, shut it down long ago in favour of hosting everything on Cloudflare These days, the developer platform has everything you could need to host 99% of apps (For transparency: I work at Cloudflare but I’ve built on them for way longer than I’ve worked for them)

• 
  Corey Edwards (@cedwardsmedia) reported 18 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.

• 
  Suat Özkaya | AI & Mobile (@suat_tw) reported 18 minutes ago

  4/6 AAIF brings competing companies to the same table. OpenAI, Anthropic and Block co-founded it, with support from Google, Microsoft, AWS, Cloudflare and Bloomberg. That is a meaningful signal: agent infrastructure is becoming too important to remain fragmented.

• 
  ILL Eagle 🇺🇲✡️ (@nh48mbk08uwvr9s) reported 18 minutes ago

  @XJosh I still don't forgive Cloudflare for what they did to shut down kiwifarms. I don't even like kiwifarms, it's a matter of principle.

• 
  Michael Tanguma (@MTanguma) reported 20 minutes ago

  AI is breaking the free internet. When a human visits a website, free access is cheap enough to subsidize. When a bot can hit it a million times for almost nothing, it isn’t. That changes the economics completely. Micropayments were part of the internet’s original vision. They never shipped, so we built the web around ads, surveillance, subscriptions, and platforms instead. AI may finally force the correction. Charge a human a few sats and they barely notice. Charge a bot a few sats every request and suddenly spam, scraping, and abuse have a real marginal cost. Cloudflare is moving this direction. Lightning makes it possible. Others will follow. The internet is finally getting a native payment layer because AI is making free access economically impossible. Full Final Settlement rip below

• 
  VC Intern (@the_vc_intern) reported 23 minutes ago

  Cloudflare is not shutting down Google Search on September 15. It is forcing Google to decide what Googlebot actually is. Cloudflare’s new controls divide crawlers into three jobs: Search, Training and Agent. Search remains allowed by default. Training and Agent traffic will be blocked on ad-supported pages for new sites and existing free customers that keep the new defaults. A crawler doing multiple jobs inherits the strictest rule. That puts Googlebot in an awkward position. Cloudflare classifies it as both Search and Training. A publisher that welcomes search indexing but rejects AI training can therefore block the entire bot. For years, bundling those jobs gave Google leverage. Publishers depended on search traffic, so refusing the crawler also meant disappearing from discovery. Cloudflare is moving that decision from a line in robots.txt to enforcement at its network edge. Google now has three choices: separate its crawlers, convince publishers to permit both uses, or lose access to pages that still want search traffic but no longer accept training as part of the bargain. The web is not dying here. Its access contract is being renegotiated. Robots.txt was a request. Cloudflare is turning it into a gate.

• 
  Tseng (@TsengSR) reported 28 minutes ago

  @iyici_ You need to learn a lot. Cloudflare is not a Domainseller. They are a Content Delivery Network with Security features for DDOS

• 
  Bobby Ivanov (@bnivanovX) reported 30 minutes ago

  @Voxyz_ai Is this essentially the cloudflare computer service attached to an agent?

• 
  Whezzel (@Whezzel) reported 30 minutes ago

  @ReinhardtOp @ReclaimTheNetHQ Tbf, there already are registries located outside the us, but they have their own challenges. The issue you'll still run into is most of the internet runs on American infrastructure, built by the us governemt and American business. Google, Amazon, Cloudflare and others. On top of that verisign manages 2 of the root dns servers, with the majority of the remaining root dns servers hosted by other American entities. To top it off, verisign distributes the zone files for the root dns servers. If Verisign decides they don't want a domain listed in dns, good luck getting to it.

• 
  Ravin Sharma (@ravinsharma7) reported 30 minutes ago

  Can attest to this. Been experimenting with many things using Cloudflare infra, the cost is literally a few dollars. In the past i read about horror stories related to cloudflare billing and support, but not sure still applies today or not. I'm just happy stuff just works without breaking the bank.

• 
  Artyom Shimanski (@a_shimanski) reported 32 minutes ago

  @truehannan @Cloudflare never got why people pay for vercel

• 
  erik@try.works (@trydotworks) reported 33 minutes ago

  @benswerd Calling cloudflare mid wtf

• 
  Abhishek Tiwari (@baanditeagle) reported 34 minutes ago

  @jacobgold It cant use the passkeys from my mac, I can’t even login on cursor or grok since cloudflare detects the VM. Is the team working on it or do we have to keep it this way?

• 
  Udit (@TherealUDP) reported 34 minutes ago

  I'm cryin dude saw a classmate flexing on story his kali linux with four terminals open using hping3 flood on a cloudflare hosted game server bragging he can down any game server/wifi he wants

• 
  Andre Gironda (@AndreGironda) reported 35 minutes ago

  @chrissanders88 Initially, sight the incoming IPs from the alert, and associated packetry, e.g., naming resolution. A good lookup service for these IPs is Spur because if a threat actor, likely coming from a vpn or proxy service, and one perhaps unique. If CloudFlare tunnel, check with abuse@

• 
  Sai Krishna ⚡️ Superblog.ai (@_skris) reported 36 minutes ago

  Can’t really fkn keep up with this company When I moved superblog (v2) from netlify to Hetzner + cloudflare, I implemented caching layer tilting towards hetzner vps because CF’s purge limits were not so great. Recently Claude told me that they give like 5 purges/second for $20/mo and 10/min for free. Seriously? I simply rewrote the caching layer with on-demand purges for superblog’s customers. Essentially the load hitting the hetzner vps is zero until a customer makes any change. Superblog will grow to a large scale even before I pay them anything. They just hate to make money I guess.

• 
  Robby Seventeen (@Robby_Seventeen) reported 41 minutes ago

  @eastdakota @Cloudflare the brand account can't post at 1am about a bad deploy. that's the whole gap right there.

• 
  Liem Do (@liemdo_) reported 44 minutes ago

  Tried out Cloudflare Containers today but it's too slow. The time it takes for containers to start up from a cold state is significantly longer compared to Google Cloud Run. I guess I still need to stick to Google Cloud Run for a while.

• 
  Maxxie wei (@xmaxxie1119) reported 46 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

• 
  Ilia Lopata (@IliaWhy) reported 47 minutes ago

  @VinodSharma10x Cloudflare turnstile …it cuts the auto bot ****…some bot spammed my inquiry form and i had to deal with that

• 
  Jim Liu (@jiahanjimliu) reported 47 minutes ago

  $CRWV: Q2 Earnings Revenue: 2.6B or 10.4B ARR EOY 2026 ARR Target: 18B-19B Backlog: 104.2B at end of Q2 (129.2B Aug 11) Adjusted EBITDA Margin: 59% Operating Margin: 5% Adjusted Net Loss: -22% Full AI Platform Contrary to misinformation on X, Coreweave serves Managed Inference, Development Tools, Orchestration and Observability and are a full AI Platform. EBITDA Margins Their EBITDA margins with software come out to 59%. $IREN's H1-4 EBITDA margins are 85% but after depreciating DC build cost for an apple-to-apple's comparison, $IREN's H1-4 EBITDA margin minus DC depreciation come out to 55%. $IREN is able to keep up on EBITDA margins without software because IREN is vertically integrated on the power and datacenter front. Coreweave's software make up for it's colocation costs to achieve 59% EBITDA margins. For comparison $NBIS has ~40% EBITDA margins. Once $IREN integrates Mirantis and DSX OS, it has a great chance of leading on EBITDA margins among the 3 Neoclouds. Backlog Coreweave hit a 129.2B backlog on earnings day Aug 11 which is a huge 25B increase from 104.2B at end of Q2. This is great news for $CRWV, $NBIS, $IREN as it shows the unprecedented demand in this sector. Financing Coreweave will likely benefit from Nvidia's 500B financing pool along with $NBIS and $IREN. However, it's high interest cost make it's Net Loss Margin -22% on what otherwise is a great inflection point of 5% operating income. In other words, Coreweave is a profitable business operations wise besides its high interest payments. This bodes well for Neocloud sector profitability as a whole. Enterprise Customers Coreweave has the widest diversification of customers among Neoclouds with: Primary Cloud: Bentley, Caterpillar, Grammarly, Isomorphic Labs, Sunday Robotics. Expanded Partnership with: Cognition, Databricks, HRT, Periodic Labs, Rescale, Runway ML. Primary cloud is important because although Coreweave and NBIS both serve Cloudflare, they are not the primary cloud for cloudflare. Likewise Shopfiy's primary cloud is GCP not NBIS. Being a primary cloud for a customer is more indicative of usage beyond of orchestrated GPUs or bare metal+k8s. Contracted Power 3.7GW by end of Q2 and 4.2GW by Aug 11. This is a majority colocation however, colocation is working for Coreweave as they still achieve 59% EBITDA margin. This sometimes results in delay but their main problem is interest expense, not colocation. $CRWV is a 55B company with 35B debt for a total of 90B EV. If $IREN can buildout 5GW and integrate in Mirantis to catch Coreweave, it has large upside as $CRWV itself still has significant upside from it's 90B EV.

• 
  Pete Stewart (@pj_stew) reported 47 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

• 
  Panyito (@CachesMR) reported 53 minutes ago

  @Vincredible123 @penisfacemcgee3 @iyici_ It's a business move, the idea is that AI crawlers will pay you to visit your site. Cloudflare can then skim some off the top for them, which is honestly not bad. Look up Cloudflare wallet

• 
  Sean (@sean_from_earth) reported 58 minutes ago

  Cloudflare is now too big and needs to be disrupted and knocked back down to size

• 
  Baraka (@zkBaraka) reported 1 hour ago

  @a_shimanski @Cloudflare Have nothing bad to say about CF, it just works. I pay them about $10/m though, I do a ton of work there it makes sense. Would have been $100 elsewhere

• 
  TopBotPicks (@TopBotPicks) reported 1 hour ago

  @Cloudflare is my registrar. My domain cannot function because my nameservers are incorrect, and I need to change them with my registrar, according to @Cloudflare. I cannot access my business email because of this. Cloudflare cannot fix this apparently. Wtff. Never buying a domain with Cloudflare again 😐