BYO-LLM

Test an LLM provider connection (owner-only) · 0 credits

Connectivity health check: resolves + decrypts the config (firm-scoped) and runs a MINIMAL invoke (a ping, short timeout) through the adapter. On success it stamps last_verified_at. Returns { ok, latency_ms, model, error?, account_issue? }. ok:true means a REAL call reached your provider with that key and model — it is not a format check. account_issue tells apart a rejection caused by YOUR account (spent balance, revoked key, model not allowed) from a connectivity failure: retrying only helps in the latter. The error message is sanitized — it NEVER echoes the api_key. No charge. Scope: mcp | * (NOT cases).

post/api/v1/llm-providers/{id}/test

Request

  • Base URL: https://legal.nexusquantum.legal
  • URL: https://legal.nexusquantum.legal/api/v1/llm-providers/{id}/test
  • Auth: HTTP bearer

Path parameters

idstring uuid required

Response

Health check result (ok=true|false, sanitized error on failure).

okboolean

true if the ping to the provider succeeded.

latency_msinteger

Latency of the minimal invoke in milliseconds.

modelstring

Model the test ran against.

errorstring

Sanitized error message (only when ok=false).

account_issue'credit_exhausted' | 'invalid_key' | 'access_denied' | 'provider_rejected'

Present when the provider ANSWERED and rejected the request because of YOUR account, not because of connectivity. credit_exhausted = the key authenticated but the balance is spent (top up with the provider; a new key will not help). invalid_key = invalid or revoked credential (replace it). access_denied = the key is valid but has no access to that model (pick a model your plan allows). provider_rejected = rejected without a classifiable reason. Retrying fixes none of them.

Changes

No recorded changes to this operation across all 1 revision of this API.