Gmail status: access issues and outage reports

Problems in the last 24 hours

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

• Website Down (37%)
• Errors (33%)
• Sign in (30%)

Live Outage Map

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

City	Problem Type	Report Time
Paris	Website Down	2 hours ago
Paris	Website Down	21 hours ago
Cannes	Sign in	1 day ago
Luc-la-Primaube	Website Down	1 day ago
Paris	Sign in	2 days ago
Medellín	Website Down	2 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.

Gmail Issues Reports

Latest outage, problems and issue reports in social media:

• 
  TeddyMan (@teddymanyeah) reported 2 minutes ago

  Do yall not know how to keep your gmail storage down? Mine isnt even 5% full and all critical emails are seperate from my shopping and social media ones.

• 
  Eyan bankulaya 😈💜 (@ZINOLEESKYTTW) reported 4 minutes ago

  @justadeniyi01 @ziccob4v Make he just use the Gmail login for android Apple Music and subscribe there it’ll reflect on the iPhone

• 
  Radu (@RamanduLight) reported 7 minutes ago

  Simple task for the spark agent in Gemini - check my spam folder for false positives. Spark: your spam folder is empty. After some back and forth: the gmail connector is not working properly. Me: 😳

• 
  Sam (@samxzk70) reported 7 minutes ago

  @ChittiDinesh @sssionggg Hello will look into a fix for that, we noticed this happened for gmail login when you should be allowed to enter instead, will push the fix soon and update. Sorry for the inconvenience 🙏

• 
  justbraceface (@murraykwamurray) reported 9 minutes ago

  @TeamYouTube I lost access to my gmail, and subsequently my YouTube channel,justbraceface. The hackers have either deleted or privated every post. I can’t access my gmail as they changed login credentials and removed 2FA and all my passkeys. I sent my DM 2 days ago. Please help

• 
  Ankit Kumar (@AnkitKu57238907) reported 12 minutes ago

  @TeamYouTube @TeamYouTube and @gmail all these steps have been completed but not able to login kindly send link on other account or send OTP on registered mobile number.

• 
  lesdoesart (@DrawsLes78493) reported 13 minutes ago

  is the identity verification tumblr a scam? Idk if it's real or scam. I have email sent me (twice) about if it's not me and it is me, l always use the website. I later press the restart password because l never created one and only login in my gmail

• 
  XZBRAT (@XZBRATIK) reported 15 minutes ago

  @ArtemisCatbat I have the same problem with epic Games support, I lost my account in 25 years and still can't regain access to it, but there's still a link to Gmail, but there's no use, there's a new mail with two-factor protection.

• 
  Parth (@Ayatoe_es) reported 16 minutes ago

  @ummNovember I use chrome too but like ill have to login my gmail n all there so too much work and im lazy 💔💔also i was device banned my insta for no reason so i cant use it on edge and chrome 💔

• 
  Maha (@maalbarki) reported 18 minutes ago

  @comet_wanderer @thisisdobleu I tried to register using gmail and FB but unfortunately it's the same issue, we need to link the account to a Thai phone number

• 
  HVAC SEO (@HVACSEO) reported 19 minutes ago

  Google is retiring the Local Services Ads dashboard! Over the next several months, every home service business running LSAs gets migrated into Google Ads as a Performance Max campaign built for pay-per-lead. The name is misleading. It does not run on YouTube, Gmail, or Display. Still Search and still Maps. Still the same ad positions. Still pay-per-lead, not pay-per-click. Still key wordless, meaning it goes off of your services on the linked google business profile What carries over: your "verified badge", your past leads, your message history, your old call recordings. If you are already verified, no insurance or license reverification What does not carry over: your performance reports. The day you migrate, the old dashboard redirects to Google Ads and your historical reports are gone. Export them before your date or lose them to the google black hole Heres what's changing Manual bidding is deprecated. Google's automated bidding sets your cost per lead now Vertical Target CPA is gone. You get one campaign-level target across every service, so if you want to bid HVAC different than plumbing you need a separate campaign for each Your weekly budget is divided by 7 into a daily average Monthly spend is capped at that daily number times 30 No more BBB callouts Add at least 6 structured callouts in the assets tab or you blend into the pack Your profile is tied to your ads tighter than ever. Change your name, address, or primary category on your Business Profile and it can pause your campaign for 24 to 48 hours while Google reviews it The timeline on all of this starts next month! Plumbing, HVAC, electrical, appliance repair, house cleaning, lawn care, roofing, pest control, and moving. Service-area and custom setups late 2026. Everyone else in 2027. You get one email 14 days out, a reminder 7 days later, and up to two weeks for performance to settle after you move. What to do before migration day -Export your Local Services Ads reports now. They do not carry over, and the dashboard is gone the day you migrate -Record your current cost per lead for each service, so you can tell if it moves after the switch -If you bid different amounts by trade, build a separate campaign for each one. A single campaign means one target across every service -Add at least 6 structured callouts in your assets tab. They replace the BBB callout you are losing. Overall, I see this as a good thing. LSA has been broken for new accounts for a while now. I'm hoping this will allow some of the newer accounts to have a chance at bidding because the last six months have not been kind to many new accounts, especially in competitive areas

• 
  Alex H (@alex_harl) reported 20 minutes ago

  @lydiahallie Improve the classifier. Mid work model switches lose in context knowledge, rely on perfect commits to not lose state, it’s tiresome having to keep mds of Fable hazards per repo (which part of codebase or tasks are not for Fable), handing off to Opus adds translation and reduces accuracy, compacts to return Fable are lossy. Classifier needs to infer from session context - long horizon potentially destructive actions are allowed for days then Fable switched on some keyword or even commits for work done by Opus*. Usage stats to show entire session not since last restart, nice to have would be /usage like codex - lifetime tokens, streaks. Don't enable by default all/never used MCPs in new CC sessions, not turn off all claude-ai MCPs but having to manually turn off individual ones like Gmail. Ability to send feedback without handing over the session - How is Claude doing do you want to share a session y/n freeform feedback no session grab. The new in-app permissions prompts are welcome and supersede a ~year of manual fudging to achieve the same, does this clone the repos to CC web, and can it be applied to older instances? We had /remote and that cloned, since last week permission prompts are in app but only new sessions. What is intended for older ones? Statusline to display what we are working on in already available text area but without us needing hooks + notes per session, you already have a recap which doesn't quite fit more work oriented (or as CC puts it a living work-coherence model). Ultraplan to not handoff to cloud repos that are not workable there, anything with secrets or session inheritance or specific permissions to systems not accessible from cloud. Native channels / messages between instances, not durable like custom MCPs (example agent mail). Cursor on Windows rendering is terrible if moved between different size screens, the entire session other than last page gets mangled as narrow text but codexes don't - sessions render properly. I don't use Cursor on Mac or move it about to test. *3 days Fable run to analyse my Databricks, ADO repos, my Az session inheritance, Databricks CLI, all pipelines, notebooks, jobs, warming up clusters and querying data - objectively a vast blast radius but that is my call. Classifier fires because some baby data engineer left a secret in a notebook 😭. How is that balanced? Equally ability to analyse own codebases for security should be a job for Fable.

• 
  Anurag Tiwari🇮🇳 (@AnuragT54755144) reported 21 minutes ago

  Hi @TeamYouTube ,I own the youtube channel @TheKiranaMart . unfortunately I forgot the gmail and phone number used to create the channel,so I can't sign in, could you please advice me on the official recovery process.thankyou.

• 
  pupupu (@0xpupupu) reported 22 minutes ago

  OPUS 5 JUST DROPPED. IT WORKS MORE AUTONOMOUSLY THAN EVER Nobody sets guardrails after the damage. They set them 30 seconds before connecting the tool, or they don't set them at all. Opus 5 launched this week built to run longer and think for itself before asking u anything. It verifies its own work and recovers from errors without stepping in for permission first. That's power. It's also more steps where it can touch something u never meant to give it access to. 1. Draft-only, always. AI never sends anything on its own, Gmail, Slack, whatever. It writes the draft, a human hits send. 1 rule, 0 exceptions. 2. No delete permission, period. Not files, not emails, not records. If a tool doesn't let u separate "write" from "delete," that's ur first red flag before u even connect it. 3. New tool, new guardrails. Don't reuse the same rules for Gmail and for a payment API. Different tool, different blast radius. 4. Ask the AI to write its own limits. Before connecting anything unfamiliar, prompt it: "walk me through what hard limits and guardrails I should consider for this specific tool." It knows its own risk surface better than most people guessing. 5. Confirm before anything irreversible runs. Sends, deletes, payments, force-pushes. If it can't be undone, it needs a human eyeball first, no exceptions for "trusted" agents. The agent isn't careless. The setup just never asked it to be careful. What's the 1 guardrail u wish u'd set up before it was too late?

• 
  AI Guides (@free_ai_guides) reported 23 minutes ago

  Paola Estefania of Better Auth spent 40 minutes at the AI Engineer World's Fair explaining why your AI agents should never act with your credentials, and made it clearer than any agent security guide I've read. This is what she walked through: 1. Give agents authority, not your credentials. Most people connect an agent to their Gmail and calendar with their own token, so it acts as them. "We need to stop giving our credentials to our agents. We need to give them authority. Stop saying pretend to be me. Start saying act for me within these limits." That one reframing, credentials to authority, is the whole protocol in a sentence. 2. Treat an agent like a new hire, not the CEO. She frames access as an org-chart question: what you hand a new employee versus what you would never hand over. "You never give someone the CEO's credentials and say go read the email. Anything could happen. With agents, that is exactly what we do." An agent should get a scoped badge of its own, the way a new hire does. 3. Capabilities beat scopes. Today you grant a broad OAuth scope like read access to an entire inbox. "A scope like read is a big thing, nothing specific. A capability is capped down. I can determine exactly what action an agent can do." That shrinks the blast radius from the whole inbox to one named action. 4. Agents need a phone book to find what they can do. Right now you wire each tool to an agent by hand, with no automatic way for it to learn what it is allowed to do. "The directory is like a phone book for the agent. Your intent gets matched to the capabilities it can actually call." Discovery becomes a lookup instead of a wiring job you maintain forever. 5. Give every agent its own identity and private key. The fix for traceability is cryptographic: each agent signs its own tokens. "Each agent gets its own private key attached to its identity. Now Gmail does not just see a user. It sees this specific agent acting on my behalf." The agent stops hiding behind you and becomes a principal you can name. 6. If you can identify an agent, you can fire it. Identity is what lets you hold an agent accountable after something goes wrong. "If the agent does something wrong, you can fire it. You revoke that one identity. Today you have to disconnect the whole thing." Revocation goes from nuking every integration to firing one bad actor. That's the identity layer for agents. Once you are wiring several of them together, one loop isn't enough, and that's graph engineering. Guide below.

• 
  Frank Jaeger (@TellTallTales) reported 23 minutes ago

  @punished_moron Is this a newer thing cause I have had gmail since like high school and never had this issue until this year.

• 
  Sonu (सोनू ) (@_sonuprakash_) reported 24 minutes ago

  Can't sign in into google account I have tried many times There is trouble with two steps verification.. @gmail

• 
  MalCare (@malcaresecurity) reported 30 minutes ago

  The first sign of a compromised site may arrive in the spam folder. Your WordPress emails worked normally for months. Then password resets, contact-form notifications and order receipts suddenly began landing in junk. It may be an authentication problem. An expired SMTP credential, overwritten DNS record or hosting migration can damage email delivery. But an unexplained reputation collapse can also be a security signal. A compromised website may send spam, host phishing pages or trigger domain warnings. From the owner’s perspective, the first visible symptom may simply be that Gmail no longer trusts anything coming from the domain. Fixing SMTP will not remove malware. Removing suspicious email content will not explain unknown outgoing messages. When deliverability changes overnight, investigate the sender—and the website producing the mail.

• 
  ‏ِ (@karoblaqx) reported 33 minutes ago

  @whizz_xD spenda wey no allow me login my acct. acct wey dey exist o e say e no exist. create new acct w same gmail e say i don use am useless app

• 
  Joseph Szobody (@jszobody) reported 37 minutes ago

  @awcodes1 What about a Chrome extension instead? That way it could actually monitor your Gmail inbox for magic login links or 2FA code and auto redirect to the link or enter the code for you. Flow ends up: click this button to receive a magic link… and a few seconds later the extension just handles it and you’re in.

• 
  Raiders (@__Raiders) reported 39 minutes ago

  add 2FAs to your Gmail before it's too later guys (probably 3FAs using your yubikeys) - more phishing/unknown login attempts incoming

• 
  BUNJEE (@BearBunjee) reported 40 minutes ago

  ChatGPT Voice just landed on desktop and it's not just talking anymore, it's actually DOING stuff. Ask it to create a thread, make a pull request, and hunt down a bug root cause in ONE voice command. that's not an assistant, that's basically a coworker. Anthropic's Claude voice mode is also out running tasks in Gmail, Slack, Notion, Canva, so the AI voice wars are officially heating up. are you actually gonna talk to your computer to code now or does that feel weird as hell to you?

• 
  SKWatcher 🇨🇦🇺🇦 (@crayzd0g) reported 40 minutes ago

  @Google @TeamYouTube Terrible support - so bad it must be deliberate. I purchased 1yr YouTube premium but used the wrong Gmail account. Can I switch the account? No you must cancel. Can I cancel? No yearly can’t cancel. Can I talk to support? No - you have none. Scrupuless😡

• 
  Raiders (@__Raiders) reported 42 minutes ago

  add 2FAs to your Gmail before it's too late guys (probably 3FAs using your yubikeys) - more phishing/unknown login attempts incoming

• 
  Prajwal Tomar (@PrajwalTomar_) reported 42 minutes ago

  I built this in a single day. Gmail inbox triage. Auto-tagged leads. Kanban pipeline. Synced to a Google Sheets lead tracker. Two years ago agencies would have quoted 2 months and $15K for this. Lovable is not a UI builder anymore. You can build literally ANYTHING inside it. I've used it almost daily for over a year. Shipped 50+ MVPs with it. For clients. For myself. For the agency. I finally wrote down the whole system. Full breakdown in the article below.

• 
  sabiu umar🛠 ENGINEER (@UmmarSabiu) reported 43 minutes ago

  @gmail Please i need help My account i can’t access it his showing couldn’t find this account And the other problem is i didn’t use my phone number Please help how can I recover it

• 
  YanXbt (@IBuzovskyi) reported 45 minutes ago

  HERMES AGENT BECOMES 10X MORE USEFUL WHEN YOU CONFIGURE THESE 5 THINGS. EACH ONE TAKES 5 MINUTES. MOST USERS NEVER TOUCH THEM. 1. THE RIGHT MODELS one model for everything = wrong model for most things. GPT-5.6 Sol: strongest reasoning. daily driver. access through your ChatGPT subscription (Plus or higher). Max plan unlocks higher reasoning effort. Grok 4.5: live X search. fastest responses. access through your X Premium+ subscription. "find me 3 high-engagement Hermes posts from the last 5 days." Grok pulls directly from X. no scraping. real-time. Kimi K3: design powerhouse. comparable quality to Claude Fable 5 at roughly 30% of the price. takes longer to generate. the quality justifies the wait. connect via Desktop app / Dashboard: Models → add provider. GPT-5.6: ChatGPT subscription → OAuth. Grok 4.5: X subscription → OAuth. Kimi K3: OpenRouter or Nous Portal. switch between them mid-session: /model [name] 2. PARALLEL TOOL CALLS Hermes used to call tools one at a time. Gmail, then calendar, then web search. sequential. now: multiple tool calls run simultaneously. "check my emails, check my calendar, tell me the weather in Dubai, and find the latest Hermes updates." four tools at once. results merge when all finish. what used to take 3 minutes takes 30 seconds. automatic after update. no config needed. hermes update 3. FASTER AND CHEAPER WEB SEARCH two improvements. one automatic, one you configure. AUTOMATIC (update only): v0.19.0 processes web pages differently. clean content straight to the agent without redundant processing steps. 60x faster. 49x cheaper. no config needed. CONFIGURE (Firecrawl): Firecrawl is the default scraping backend. strips HTML, ads, navigation, scripts. returns only the text your agent needs. 500 free credits per month on free tier. get your key from firecrawl .dev. add to .env: FIRECRAWL_API_KEY=your_key Nous Portal subscribers: Firecrawl is included through Tool Gateway. no separate key needed. SAVE MORE (auxiliary model): web summarization defaults to your main model. route it to a cheap model: auxiliary: web_extract: model: google/gemini-3-flash-preview cheap model reads the page. premium model reasons about the content. 4. MORNING BRIEF WITH EMAIL + CALENDAR connect Gmail and Google Calendar via MCP: 1. go to mcp .zapier.com 2. add Gmail: enable read and draft only. never enable send. one automated email from the wrong context can cost a relationship. 3. add Google Calendar: read access. 4. click connect → sign in → regenerate token 5. paste the token into Hermes chat tell your agent: "create a

• 
  Matt (@m13v_) reported 47 minutes ago

  the fix was never a smarter model. it's that runner runs on the desktop with 42 connectors wired in, so gmail, linear and hubspot are one context instead of me tabbing between four windows to hand-carry the numbers.

• 
  Tori Knudsen | UGC Creator (@ToriKnudsenUGC) reported 52 minutes ago

  @naturallyshannn My credit is frozen so im not worried about things being stolen. And you can give you a VA a login to gmail without sharing your password.

• 
  gust (@gust_wd) reported 53 minutes ago

  How do you let Hermes agents use Gmail & Stripe without giving them API keys? A common setup is to put credentials in .env files. But Hermes agents read files, call tools, summarize logs, and debug errors. So the secret you thought was “safe” can leak into model context and become accessible to the wrong people. OneCLI changes this. You don’t give the credential to Hermes. The secret stays outside the agent. Hermes gets access. The model doesn’t need to see the key.