Checking…
Fetching
/api/healthGateway (usatechsolution.ai)
—
Database
—
Upstream models (Azure AI Foundry)
—
PayPal checkout
—
What is monitored
The health endpoint runs a database round-trip and a one-token embeddings request against the upstream every 60 seconds at most (results are cached). It is public: GET https://usatechsolution.ai/api/health. Use it from your own monitors; status is ok, degraded or down (HTTP 503 when down).
…
Limits and guarantees
- Default per-key limits: 60 requests/min, 8,192 output tokens per request, $20/day spend cap (token keys). Raise them by email.
- Streaming responses run for up to 60 seconds per request.
- Nothing is billed for requests the upstream rejects (HTTP 502
upstream_error). - Incidents and planned changes are announced on the dashboard. Report problems to anylohar49@gmail.com with the model id and time of the request.