Refreshed every by the Worker
- Live models
- Checked
- Average
- Next check
Entry point
- Base URL
- Endpoint
- /chat/completions OpenAI-compatible
Model health
What each light means
- Up
- Responding normally. The number beside it is the response time.
- Degraded
- Still responding, but took 8 seconds. Do not rely on it for important answers.
- Limited
- Responding, but the token budget ran out mid-reasoning so no answer came through. Raise max_tokens in the cron in workers.js.
- Non-chat
- Listed, but does not accept chat requests. Shown as-is, without being forced into either up or down.