ModelsCompareBest forBenchmarksStatusPricingAPI
Status / DeepSeek

Is DeepSeek down?

All systems operational

DeepSeek's official status feed, polled every 5 minutes. Direct API latency probes are not enabled for DeepSeek yet. Last checked 2026-09-18 03:40 UTC.

Current status
operational
checked 3 min ago
Incident days (90d)
1
days inside an official incident window
Incidents (30d)
0
0 in 90 days · 0 stored
Observed uptime (30d)
100%
sampled availability
Last incident
—
none on the feed
History

90-day status history

90 days agoToday
OperationalDegradedOutageNo data
Official feed

Incident history

status.deepseek.com ↗

No incidents on DeepSeek's official feed in the window we retain.

Titles and impact levels are DeepSeek's own classifications, mirrored from the official feed. Consumer-product incidents appear alongside API incidents when the provider reports them on the same feed.
Right now

Other providers

Full board →
OpenAI95.3079%Anthropic100%Google100%Mistral AI100%Cohere100%Moonshot AI100%MiniMax100%Perplexity100%NVIDIA100%Meta—xAI—Qwen—Amazon—Microsoft—Z.ai—Baidu—Xiaomi MiMo—ByteDance Seed—StepFun—Meituan LongCat—Tencent Hunyuan—Upstage—

Is DeepSeek down right now?

DeepSeek is currently reported as: All systems operational. DeepSeek's official status feed, polled every 5 minutes. Direct API latency probes are not enabled for DeepSeek yet. Last checked 2026-09-18 03:40 UTC.

How does llmmetric know whether DeepSeek is down?

llmmetric polls DeepSeek's official status feed every five minutes and records the reported state in daily buckets. The 30-day uptime figure is sampled availability, while official incident windows separately mark affected days in the 90-day history.

What can I do while DeepSeek is down?

If DeepSeek is down, a gateway that routes across providers with automatic failover keeps your app up. Respan does this with one API key across every provider on this site.

Built by Respan
DeepSeek down and your app is stuck?

Respan's gateway routes across providers with automatic failover, so an outage on one becomes a retry on another. One API key, full tracing.

Set up failover freerespan.ai ↗