Is Hugging Face down?

What works right now

• ●UpHub APIYes: huggingface.co/api answers.200 in 218 ms
• ●UpFile downloadsYes: model files resolve for download.307 in 217 ms

Failing for you? Check whether it is you

1. Is it you?

   A 401 or 403 on one repo is usually a gated model or a missing token, not an outage.

What to do while Hugging Face is down

When Hugging Face is down, model and dataset downloads, the Hub API, Spaces or Inference Endpoints may fail. Models you have already downloaded load from your local cache.

1. Load from your cache

   Offline mode makes transformers, diffusers and huggingface_hub use only ~/.cache/huggingface/hub.

   export HF_HUB_OFFLINE=1

   Copy

2. Download ahead of time

   Pull the whole repo into the cache while the Hub is up.

   huggingface-cli download Qwen/Qwen2.5-0.5B-Instruct
   # newer releases also ship it as: hf download REPO

   Copy

3. Use a mirror

   hf-mirror.com is a community mirror, not run by Hugging Face. Use it for public models only.

   export HF_ENDPOINT=https://hf-mirror.com

   Copy

Last 24 hours

Last 30 days

Incident timeline

Outages our checks saw, and incidents from the official status page, newest first.

1. OfficialOct 7, 7:33 UTC · 5 hScheduled maintenance: Inference Endpoints (H200)

   Maintenance ended

2. OfficialSep 24, 18:00 UTC · 15.9 hSome GPU Jobs and Spaces failing to start

   The issue is resolved. If your job failed between 18:00 UTC on September 24 and 09:10 UTC on September 25, please resubmit it. If your Space is still showing a Runtime error, restart it from the Space settings.

3. OfficialSep 22, 18:48 UTC · 3 daysSlower downloads in the Asia-Pacific region

   The upstream network capacity issue affecting our Singapore CDN location has been resolved, and download performance for Asia-Pacific users is back to normal. We've returned traffic routing to its usual configuration. Thank you for your patience.

4. OfficialSep 16, 7:00 UTC · 2 hScheduled maintenance: Hub

   Maintenance ended

5. OfficialSep 2, 7:00 UTC · 1 hScheduled maintenance: Hub

   Maintenance ended

6. OfficialAug 17, 7:00 UTC · 2 hScheduled maintenance: Spaces

   Maintenance ended

7. OfficialAug 7, 15:14 UTC · 8 hElevated error rate - AWS CDN (Singapore)

   Our fix is now in place and this incident is resolved. Applying it caused one final short period of failed downloads between 23:00 and 23:14 UTC, as we noted could happen. Downloads from the Singapore region have been stable since, and we have added capacity to prevent this from

Official status

The official page: https://status.huggingface.co/.

Get alerts, or show it in your README

Follow the RSS feed of Hugging Face incidents in any reader, or paste this badge into a README; it updates itself.

[![Hugging Face status](https://tensorust-status.vercel.app/api/badge?id=hugging-face)](https://tensorust.com/status/hugging-face/)

How we check Hugging Face

Every 5 minutes, from Washington, D.C., with no API key. A failed check is retried once after 2 seconds before it counts; Hugging Face is down when every critical check fails, and degraded when some do or the official page reports a problem.

• https://huggingface.co/api/models?limit=1: huggingface.co/api (search, metadata).
• https://huggingface.co/openai-community/gpt2/resolve/main/config.json: What from_pretrained and hf download hit first.

Other services