Bluesky outage locks out web users as mobile apps still work

If Bluesky won’t load or your timeline is frozen, you are not alone. A surge of user reports indicate that the social platform is suffering from a partial outage, with loads of people who are locked out on the web while its mobile app still seems to work for some.

Downdetector, a third-party service that monitors internet outages, logged thousands of problem reports for Bluesky in an alarming spike — lending credibility to the idea of an issue affecting more people than simply a few customers. However, Bluesky’s own status dashboard currently — and always — shows all systems as operational, an inconsistency that can occasionally occur in fast-moving or partial incidents.

What’s happening now with Bluesky’s partial web outage

Depending on the user, symptoms vary and include timelines not refreshing, profile pages not loading, and error messages appearing in web views. Some users have said the Android and iOS apps are still able to connect, although posting and search may be intermittent. That split — mobile working while the web stalls — is often an edge or front-end issue rather than a total platform crash.

Bluesky is built atop the AT Protocol, a federated system intended to be able to scale through many services. In reality, that could mean various parts (a web front end, APIs, personal data server and authentication) get into trouble on their own. A hiccup in a content delivery network, a DNS change that hasn’t propagated evenly, or the little web build behaving badly — all can make the browser experience bad when the app, which talks to APIs more directly, is there already limping.

Why The App Might Succeed Where The Web Has Failed

Mobile apps often cache credentials, and as long as the API endpoints are not removed, can remain operational even if the website is having issues. The web client, however, is reliant on new assets that are loaded in via CDNs, strong browser security policies, and cookie-based authentication. There’s a CDN misconfiguration, an expired certificate, or a bug in the latest deployment of the web app — and the breaking point is sometimes reached first (and loudly) in the browser.

Another potential would be the rate limiting or an influx of traffic that is getting routed to this specific cluster. We have seen the same phenomenon in other social platforms, with the website timing out but phone apps continuing to serve cached timelines and gradually deliver new posts once capacity re-establishes.

How to know it’s not just you during a Bluesky outage

• Consult third-party outage trackers like Downdetector for reports of spikes related to Bluesky.
• Check the platform’s status page to see if they acknowledge an incident. Be aware that partial outages can take time to show up.
• Scan app-wide conversations within the platform; members usually report when they can access and what they’re able to do (post, DMs, search).
• Check broader internet telemetry, such as Cloudflare Radar, to see whether the issue is platform-specific or related to regional connectivity.

Things to try on your own to restore Bluesky access

• Switch networks: switch to or from Wi‑Fi or cellular to eliminate local routing issues.
• Temporarily disable VPNs or ad blockers; some can disrupt CDN assets and auth flows.
• On the web, hard refresh and clear site data (cache + cookies) for Bluesky. Force stop and restart the app; try not to tap Post several times to avoid submitting multiple duplicates when service is restored.
• If you manage to get into the app, don’t touch that “logout” button; when authentication services are wonky, it may be tougher getting back on than simply staying logged in.

Context from past social outages on major platforms

Massive social networks often encounter short, high-impact outages as they grow. X struggled with rate limiting that temporarily took down timelines for lots of users, while image uploads and feeds have failed on photo-first apps under traffic surges. Federated services like Mastodon have also experienced instance-specific slowdowns that seem like site-wide outages to a user.

Bluesky, which has expanded to millions of accounts since it went public, has experienced the same sort of jump toward heavier peaks and more complex routing. A status page may show green in the first few minutes of an incident or when the problem is restricted to a certain surface, like the web client.

What to expect next as Bluesky engineers resolve issues

Partial outages like this generally correct themselves as soon as engineers roll back a bad change, reconfigure edge caching, or add capacity. Once service is restored, note that queued actions may be pushed out of order and drafts created during errors might not persist.

For now, if the web isn’t working for you, try using the mobile app or keep an eye on independent outage monitors, and hold off on any major account changes. If previous incidents tracked by monitoring companies are any indication, over the next day, service stability should only continue to improve as mitigation is deployed and global caches update.