Check your DNS
Flush your DNS cache: ipconfig /flushdns on Windows, sudo dscacheutil -flushcache on macOS.
Site web · chatgpt.com
ChatGPT est-il en panne ?
chatgpt.com
ai
OPÉRATIONNEL
Checked 1s ago
4/4 regions returned data
every 5 min · HTTP probe
direct probe, not crowd-reported
Résultats par région
US East
OPÉRATIONNEL
22 ms
HTTP 403
US West
OPÉRATIONNEL
25 ms
HTTP 403
Europe
OPÉRATIONNEL
23 ms
HTTP 403
Asia
OPÉRATIONNEL
33 ms
HTTP 403
Détails HTTP
- HTTP status
- 403
- Response time
- 22 ms
- Region consensus
- 4/4 reachable
- 24h success rate
- 50% (1/2 probes)
- 24h latency
- 26–26 ms · 26 ms avg
- Last successful probe
- 1s ago
Décomposition des temps
- DNS lookup
- 9 ms
- TCP connect
- 1 ms
- TLS handshake
- 7 ms
- First byte
- 5 ms
- Total
- 22 ms
Réseau
- Host
- chatgpt.com
- Port
- 443
- Transport
- HTTPS
Que signifie ce résultat
ChatGPT is reachable from all 4 monitoring regions.
If you still can't connect, the outage is on your side: DNS cache, VPN routing, ISP filtering, or a browser-level block. Try a different network or device to confirm.
Que faire ensuite
Try another network
Switch between Wi-Fi and mobile data, or disable your VPN. If it works there, the issue is on your primary network.
Browser reset
Try an incognito window or a different browser — a bad extension or cached error page can masquerade as downtime.
FAQ
Why is ChatGPT down right now?
Outages typically come from CDN/DNS issues, failed deploys, regional routing problems, or DDoS. We can't see ChatGPT's internal cause — we only confirm reachability from each monitoring region. If all 4 regions failed, it's a confirmed outage on ChatGPT's side. If only some failed, the impact is regional.
Is ChatGPT down for everyone or just me?
Look at the per-region grid above. If every region succeeded, the issue is local to your network — flush DNS, switch networks, or disable your VPN. If two or more regions failed, the outage is on ChatGPT's side.
How often does isitdown.io re-check ChatGPT?
Every 5 minutes via background HTTP probes from US East, US West, Europe, and Asia. You can also re-run on demand by reloading — the verdict above reflects the latest probe.
Voir aussi · ai
Méthodologie
- Sonde
- We send a GET request to the target URL from each region and time the full response.
- Compté comme EN PANNE si
- Every region fails the request (connect refused, DNS error, timeout, or HTTP 5xx).
- Compté comme DÉGRADÉ si
- At least one region succeeds but at least one fails — the outage is partial or regional.
- Détail
- Per-region results include HTTP status code, response time, and best-effort DNS / connect / TLS / first-byte phase timings when the Node HTTP client exposes them.
- Cadence
- Every 5 minutes, in parallel across 4 monitoring regions (US East Virginia, US West Oregon, Europe London, Asia Singapore).
- Cibles limitées en débit
- If a host returns 429 or consistently drops connections from our IPs, we cap retries at 3 and report the last observed status — we do not flood the target to confirm the outage.
- Source des données
- Direct probes from our monitoring infrastructure. We do not aggregate crowd reports, Twitter mentions, or DownDetector signals — every result on this page is a live network request.