Omniplex
Affected components
Issues have been resolved and services are stable
A fix has been applied and we are monitoring the results
We are investigating an outage with our proxy service
At approximately 1:45 AM MST on August 22, 2026, a deploy to our Discord API proxy (byte-proxy) briefly took down every Omniplex service that talks to Discord through it, bots, the website, and the API. The outage lasted roughly 3 minutes before our team identified and patched the issue.
All services were affected for the duration:
No data was lost. The impact was limited to request failures during the outage window, everything recovered on its own once the patch went out.
The deploy rewrote how the proxy rate-limits and authenticates outbound Discord requests, moving from a single shared limit to per-credential tracking, and finishing support for callers supplying their own bot tokens instead of always using the proxy's shared one. A bug in that change caused Discord requests routed through the proxy to fail or hang instead of completing, which meant everything depending on the proxy for Discord access went down with it at the same time.
The team patched the newly-rewritten Discord rate-limiting logic and finished the caller-provided-token authentication path it depended on. Once deployed, Discord requests through the proxy resumed processing normally and every affected service recovered without further action.