Server Settings
Platform mail, app, and LLM endpoint configuration
App & Mail
Checking Postmark status…
Update transactional email and public URL settings here. The token is never shown back once saved.
Signed-up Users
| User | Plan | Status | Role | Verified | Last Sign-in | Sessions | Created | Namespace | Save |
|---|---|---|---|---|---|---|---|---|---|
| Loading… | |||||||||
Support Requests
| Category | Subject | Status | Priority | Created | |
|---|---|---|---|---|---|
| Loading… | |||||
LLM Pool Status
--
Total
--
Enabled
--
Healthy
--
Active Endpoint
LLM Endpoints
Loading...
Priority Guide
Lower priority number = tried first. Set local endpoints (vLLM, Ollama over WireGuard) to low numbers (10-30)
and cloud APIs (OpenAI, Grok) to higher numbers (50-100) as fallbacks.
429 handling: If an endpoint returns 429 (rate limited), the system waits and retries on the same endpoint — it won't skip to a paid cloud fallback just because a local endpoint is busy.
5xx handling: Server errors retry once, then fall through to the next endpoint. Connection failures skip immediately.
429 handling: If an endpoint returns 429 (rate limited), the system waits and retries on the same endpoint — it won't skip to a paid cloud fallback just because a local endpoint is busy.
5xx handling: Server errors retry once, then fall through to the next endpoint. Connection failures skip immediately.