Wrong Rate Limts!
Hi, I’m having an issue where OpenClaw keeps failing with:
Provider ... is in cooldown (all profiles unavailable) (rate_limit)
Initially OpenAI Codex hit its rate limit, but OpenClaw continued working using Anthropic without problems. Later Anthropic also started returning rate_limit, and now both providers are marked as in cooldown, so every request fails immediately.
However, my actual usage is very low (about $4 total usage, far below limits), and Anthropic shows no token rate limit spikes. Requests are also infrequent (~13k tokens every 30 minutes with caching).
I already restarted the gateway, but the cooldown persists right after startup.
Has anyone seen OpenClaw keep providers in cooldown even after restart, or misclassify errors as rate_limit?
Provider ... is in cooldown (all profiles unavailable) (rate_limit)
Initially OpenAI Codex hit its rate limit, but OpenClaw continued working using Anthropic without problems. Later Anthropic also started returning rate_limit, and now both providers are marked as in cooldown, so every request fails immediately.
However, my actual usage is very low (about $4 total usage, far below limits), and Anthropic shows no token rate limit spikes. Requests are also infrequent (~13k tokens every 30 minutes with caching).
I already restarted the gateway, but the cooldown persists right after startup.
Has anyone seen OpenClaw keep providers in cooldown even after restart, or misclassify errors as rate_limit?
The AI that actually does things. Emails, calendar, home automation - all from your favorite chat app. New shell, ready to help. The lobster way.
134,827Members
Resources
Recent Announcements
Similar Threads
Was this page helpful?
