Is Modal down right now?
HTTP endpoint monitoring - checked every 5 minutes · How we classify outages
Modal is currently down - official outage reported - 644ms HTTP response. Last checked . 90-day uptime: 99.3%.
Storage refactor following AWS us-east-1c issues
Stay informed
Follow @tickerr_ai
We post when tier 1 LLM APIs go down - before the official status page updates.
Follow on X →Or get the weekly reliability digest:
Add a status badge
Show live Modal status on your site, README, or docs.
90-day uptime
99.3%
18 incidents (90d)
Response now
644ms
HTTP p50 (7d)
864ms
median ping response
HTTP p95 (7d)
2921ms
tail ping response
Response time (7 days)
p50 864ms·p95 2921msⓘ HTTP response times from Tickerr's monitoring server to Modal's status or homepage endpoint - not API inference latency or time-to-first-token (TTFT). For API benchmark data, see third-party sources like Artificial Analysis.
90-day uptime
Incident history
CPU functions, Snapshot restores (beta), and Image builds went down.
Following mitigations made to our storage layer, we are going to move some data around during off-peak hours. Existing runs should mostly proceed without disruption, but scheduling will temporarily be…
Due to an H200 capacity shortage, we are seeing degraded H200 scheduling latency
Due to a code change, CIDR allowlists were applied too aggressively, and egress used for some sandbox functionality (including tunnels, exec API, filesystem API, and connection tokens) was unintention…
Due to a DNS change, container scheduling was degraded
Function execution is currently degrated
Functions running in us-west input plane are down.
We're noticing increased latency for web endpoints affecting some requests
A small percentage of Function containers failing to start up
Modal.com web dashboard degraded. Metrics are slower to load.
Some web endpoint requests that are proxied through our us-east input plane are failing.
Sandboxes are experiencing transient errors in scheduling.
We're investigating an issue where function call, container scheduling, and web endpoint latencies are degraded.
Sandbox scheduling was unavailable from 3:58 – 4:06 UTC
Task logs have been delayed since 2:46 PM ET
Related pages
About Modal status
Modal is a cloud compute platform for running Python workloads including ML inference, data pipelines, and AI agents. Modal downtime affects function execution and deployment. Modal is commonly used for custom model hosting and batch AI workloads.