---
title: "AI Recommendation Index — measurement disclosure"
method: GET
path: "/api/index/methodology"
tags: ["AI Index"]
---

# AI Recommendation Index — measurement disclosure

`GET /api/index/methodology`

## Request

- Base URL: `https://api.orbator.io`
- URL: `https://api.orbator.io/api/index/methodology`
- Auth: one of:
  - none
  - API key in header `X-Api-Key`
  - HTTP bearer

What the Index actually measures, read live from the collectors: the effective model per engine (env override or default), sampling cadence, the 28-day window and minimum-sample gate, current extraction version, and the trend comparability guardrails. Backs the human methodology page at https://www.orbator.io/ai-index/methodology

## Response `200`

Measurement parameters currently in effect

- IndexMethodology
  - `source` string
  - `license` string
  - `methodology_url` string
  - `window` object
    - `weeks` integer
    - `days` integer
  - `min_sample` integer: Runs required in the window before a category publishes
  - `sampling_cadence` string
  - `extraction_version` integer: Current extraction pipeline version stamped on stored sightings
  - `sampling` object: Repeated-sampling design currently in effect (samples per prompt-engine pair per cycle)
    - `samples_per_prompt` integer: Global N per prompt-engine pair; 1 = single-draw
    - `per_engine` object: Effective N per engine (env override or engine default)
    - `sampled_categories_limit` integer: Only the top-N categories by exposure sample at N>1
    - `customer_samples_per_prompt` integer: N for customer-lane visibility runs
  - `engines` object[]
    - `id` string
    - `model` string: Effective measured model (env override or default)
    - `available` boolean: API key configured — the engine is collecting
    - `era_filter` string, nullable: Model prefix public aggregation is restricted to during a changeover; null when none
  - `trend_guardrails` object: Windows must be comparable before a trend publishes (else suppressed)
    - `min_engine_runs` integer: Minimum runs per configured engine in both windows
    - `low_engine_runs` integer: Below this per-engine run count a passing trend is graded low confidence
    - `volume_ratio` number[]: Allowed current/prior run-volume ratio range
    - `flat_band_pp` number: Share deltas within ± this band report as flat
    - `small_sample_flat_band_pp` number: Widened flat band under small samples
    - `small_sample_runs` integer: Sample size below which the widened band applies
    - `confidence_levels` string[]

## Other responses

- `429`: Hourly quota exceeded for the current tier. Retry after the Retry-After header.

---

[API](https://skmtc.dev/orbator/apis/orbator-public-api.md) · [All operations](https://skmtc.dev/orbator/apis/orbator-public-api/llms.txt) · [OpenAPI document](https://skmtc.dev/orbator/apis/orbator-public-api/revisions/7d1cf20c2163?raw)