Premium Probe Series
1 creditGET /api/premium/probe/seriesDaily aggregates from the active LLM probe (15-min cycle). Per-provider TTFB, total latency, status mix, and budget consumption rolled up per UTC day.
When to use this endpoint
For SLA reports, vendor reliability comparisons, or post-incident timeline reconstruction. Companion to the free /api/probe/latest 24-hour summary.
Parameters
| Name | In | Type | Description |
|---|---|---|---|
| provider* | query | string | Provider key (anthropic, openai, gemini, etc.)e.g. anthropic |
| from | query | string | Start date (default 30d ago)e.g. 2026-04-09 |
| to | query | string | End datee.g. 2026-05-09 |
* required
Example response
{ "ok": true, "provider": "anthropic", "series": [ { "date": "2026-05-09", "p50_total_ms": 1234, "p95_total_ms": 3120, "error_rate_pct": 0.4, "samples": 96 } ], "billing": { "credits_charged": 1, "credits_remaining": 49 } }Code samples
Python SDK
tf._get("/premium/probe/series", provider="anthropic")TypeScript SDK
await fetch("https://tensorfeed.ai/api/premium/probe/series?provider=anthropic", { headers: { Authorization: "Bearer tf_live_..." } });FAQ
What is the budget cap?
Each probe invocation has a per-provider daily budget; once exhausted, samples drop to zero for that day. Budget consumption is reported alongside the latency stats.