Inference.net status

Q: Is Inference.net down right now?

Inference.net does not publish a machine-readable status feed. This page shows the success rate LLM Gateway measured on real Inference.net traffic over the last 24 hours and 90 days.

Q: How is Inference.net uptime measured?

Uptime is the share of requests to Inference.net through LLM Gateway that completed without a provider or gateway error. Invalid requests (4xx caused by the request) and traffic on customers' own provider keys are excluded.

Q: What happens to my requests when Inference.net has an outage?

LLM Gateway retries failed requests on another healthy provider serving the same model and routes new traffic away from degraded providers. 5 of the 7 Inference.net models on LLM Gateway can fail over this way.

Q: How often does this page update?

Every five minutes. Measured uptime comes from LLM Gateway's hourly traffic rollups; the official status and incidents come from the provider's status page.

Is Inference.net down? Live official status, incidents, and the uptime LLM Gateway measured on real Inference.net traffic.

No official feed

Measured uptime

Success rate of Inference.net requests through LLM Gateway, by UTC day.

Last 24 hours

97.48%

1,431 requests

Last 90 days

99.58%

200,582 requests

Aug 11: Not enough traffic to measure
Aug 12: Not enough traffic to measure
Aug 13: Not enough traffic to measure
Aug 14: Not enough traffic to measure
Aug 15: Not enough traffic to measure
Aug 16: Not enough traffic to measure
Aug 17: Not enough traffic to measure
Aug 18: Not enough traffic to measure
Aug 19: Not enough traffic to measure
Aug 20: Not enough traffic to measure
Aug 21: Not enough traffic to measure
Aug 22: Not enough traffic to measure
Aug 23: Not enough traffic to measure
Aug 24: Not enough traffic to measure
Aug 25: Not enough traffic to measure
Aug 26: Not enough traffic to measure
Aug 27: Not enough traffic to measure
Aug 28: Not enough traffic to measure
Aug 29: Not enough traffic to measure
Aug 30: Not enough traffic to measure
Aug 31: Not enough traffic to measure
Sep 1: Not enough traffic to measure
Sep 2: Not enough traffic to measure
Sep 3: Not enough traffic to measure
Sep 4: Not enough traffic to measure
Sep 5: Not enough traffic to measure
Sep 6: Not enough traffic to measure
Sep 7: Not enough traffic to measure
Sep 8: Not enough traffic to measure
Sep 9: Not enough traffic to measure
Sep 10: Not enough traffic to measure
Sep 11: Not enough traffic to measure
Sep 12: Not enough traffic to measure
Sep 13: Not enough traffic to measure
Sep 14: Not enough traffic to measure
Sep 15: Not enough traffic to measure
Sep 16: Not enough traffic to measure
Sep 17: Not enough traffic to measure
Sep 18: Not enough traffic to measure
Sep 19: Not enough traffic to measure
Sep 20: Not enough traffic to measure
Sep 21: Not enough traffic to measure
Sep 22: Not enough traffic to measure
Sep 23: Not enough traffic to measure
Sep 24: 92.64% uptime over 693 requests
Sep 25: Not enough traffic to measure
Sep 26: Not enough traffic to measure
Sep 27: Not enough traffic to measure
Sep 28: 0.00% uptime over 103 requests
Sep 29: Not enough traffic to measure
Sep 30: 98.33% uptime over 958 requests
Oct 1: Not enough traffic to measure
Oct 2: 99.94% uptime over 1,710 requests
Oct 3: 98.57% uptime over 210 requests
Oct 4: 100% uptime over 1,058 requests
Oct 5: Not enough traffic to measure
Oct 6: 99.32% uptime over 4,993 requests
Oct 7: 99.63% uptime over 17,992 requests
Oct 8: 99.42% uptime over 8,924 requests
Oct 9: 99.66% uptime over 5,244 requests
Oct 10: 99.56% uptime over 4,732 requests
Oct 11: 100% uptime over 1,958 requests
Oct 12: 99.97% uptime over 9,703 requests
Oct 13: 99.81% uptime over 5,742 requests
Oct 14: 99.64% uptime over 5,611 requests
Oct 15: 99.97% uptime over 8,989 requests
Oct 16: 99.89% uptime over 7,197 requests
Oct 17: 100% uptime over 3,404 requests
Oct 18: 99.96% uptime over 4,449 requests
Oct 19: 99.94% uptime over 8,760 requests
Oct 20: 99.95% uptime over 3,705 requests
Oct 21: 99.82% uptime over 6,132 requests
Oct 22: 100% uptime over 1,569 requests
Oct 23: 99.97% uptime over 3,877 requests
Oct 24: 99.99% uptime over 7,899 requests
Oct 25: 100% uptime over 12,306 requests
Oct 26: 100% uptime over 2,318 requests
Oct 27: 99.93% uptime over 8,544 requests
Oct 28: 99.92% uptime over 10,735 requests
Oct 29: 99.72% uptime over 7,420 requests
Oct 30: 99.90% uptime over 4,182 requests
Oct 31: 99.92% uptime over 6,361 requests
Nov 1: 100% uptime over 2,586 requests
Nov 2: 100% uptime over 2,706 requests
Nov 3: 100% uptime over 3,747 requests
Nov 4: 100% uptime over 1,415 requests
Nov 5: 100% uptime over 1,001 requests
Nov 6: 99.99% uptime over 8,702 requests
Nov 7: 100% uptime over 1,620 requests
Nov 8: 95.75% uptime over 847 requests

30 days ago90 days agoToday

≥ 99.5%
≥ 98%
≥ 95%
≥ 90%
< 90%
No data
Official incident

Automatic failover

When Inference.net fails or slows down, LLM Gateway retries the request on another healthy provider serving the same model, with no change to your code. 5 of 7 Inference.net models can fail over.

Inference.net models and pricing

Frequently asked questions

Is Inference.net down right now?

Inference.net does not publish a machine-readable status feed. This page shows the success rate LLM Gateway measured on real Inference.net traffic over the last 24 hours and 90 days.

How is Inference.net uptime measured?

Uptime is the share of requests to Inference.net through LLM Gateway that completed without a provider or gateway error. Invalid requests (4xx caused by the request) and traffic on customers' own provider keys are excluded.

What happens to my requests when Inference.net has an outage?

LLM Gateway retries failed requests on another healthy provider serving the same model and routes new traffic away from degraded providers. 5 of the 7 Inference.net models on LLM Gateway can fail over this way.

How often does this page update?

Every five minutes. Measured uptime comes from LLM Gateway's hourly traffic rollups; the official status and incidents come from the provider's status page.