Trigger a voice-eval run
Trigger an evaluation run against a golden set. The runner replays every case (single-turn topic match or multi-turn persona simulation) against the target agent, then grades transcripts with an LLM judge. Optionally override the STT/TTS provider or LLM model for A/B comparisons. Returns the run id plus pass/fail/latency totals and an is_simulated flag when STT/TTS latencies are stubbed rather than measured.
Authorizations
Dashboard JWT token from Clerk
Headers
Stripe-style idempotency token. Pass a stable, client-generated value (1-255 chars) to dedupe retries on transient timeouts. The same key+credential+path replays the original response for 24h on 2xx (5min on 4xx, 30s on 5xx). Returns 409 if a concurrent request with the same key is already in flight; replayed responses include the Idempotency-Replay: true response header.
1 - 255Sandbox opt-in for Clerk-session-authenticated requests. Set to true to route the call through the test-mode pipeline: no real provider delivery, no credits deducted, response meta.test_mode: true. Ignored for live API keys (dv_live_sk_*) — server-to-server clients must use a test-prefixed key (dv_test_sk_*) to exercise sandbox. Test-prefixed keys unconditionally enable sandbox regardless of this header.
true, false Body
Golden set to execute.
200Voice agent under test (omit for provider-only A/B runs).
200Specific agent prompt version to evaluate.
200Trigger origin (defaults to manual).
manual, pre-deploy-ci, scheduled STT provider override (A/B testing).
100TTS provider override (A/B testing).
100LLM model override (A/B testing).
200