Research API — v1

MindFrame Research API

Anonymized, aggregate cognitive data for researchers and academics. Access Hive-level insights — calibration patterns, bias rates, and reasoning quality — without exposing individual user data.

Use Cases

Cognitive Science Research

Study metacognition patterns at scale — calibration curves, confidence biases, and self-awareness across demographics.

Bias Studies

Access aggregate bias prevalence data across 15 challenge modes. Identify which biases are most persistent across populations.

Calibration Research

Investigate how confidence and accuracy relate — test theories about overconfidence, the Dunning-Kruger effect, and training interventions.

Authentication

The Research API requires an approved API key. Keys are issued only to verified academic institutions and research organizations. All requests must include the key in the Authorization header.

GET /api/v1/research/hive-overview
Authorization: Bearer YOUR_RESEARCH_API_KEY
Content-Type: application/json

Request an API key

Send your institutional affiliation, research purpose, and intended use to:

research@usemindframe.com →

Endpoints

GET/api/v1/research/hive-overview

Returns aggregate statistics across all MindFrame users. All data is anonymized and aggregated — no individual user data is exposed.

Example Response

{
  "totalMinds": 4821,
  "avgCalibrationError": 18.4,
  "avgCompositeScore": 67.2,
  "overconfidenceRate": 0.61,
  "underconfidenceRate": 0.22,
  "wellCalibratedRate": 0.17,
  "sessionCountLast30Days": 12047,
  "mostCommonBias": "confirmation_bias",
  "modeBreakdown": {
    "bias-hunter": { "avgScore": 64.1, "sessions": 3821 },
    "calibration-lab": { "avgScore": 71.3, "sessions": 2904 }
  }
}

GET/api/v1/research/calibration-insights

Detailed calibration data showing how confidence relates to accuracy across sessions. Includes Brier scores, overconfidence gaps, and calibration curves by mode.

Parameters

modestring?Filter by training mode slug (e.g. bias-hunter)

periodstring?Time period: 7d | 30d | 90d | all (default: 30d)

Example Response

{
  "period": "30d",
  "mode": null,
  "avgBrierScore": 0.21,
  "avgCalibrationError": 18.4,
  "avgOverconfidenceGap": 12.1,
  "calibrationCurve": [
    { "confidenceBucket": "0-10", "actualAccuracy": 0.08, "count": 412 },
    { "confidenceBucket": "40-50", "actualAccuracy": 0.43, "count": 1821 },
    { "confidenceBucket": "90-100", "actualAccuracy": 0.71, "count": 934 }
  ],
  "trendsOverTime": [
    { "week": "2026-W01", "avgCalibrationError": 21.3 },
    { "week": "2026-W08", "avgCalibrationError": 17.9 }
  ]
}

Rate Limits & Terms

Rate Limits

1,000 requests / day per API key
100 requests / hour burst limit
Responses cached for 15 minutes

Data Terms

All data is anonymized and aggregated
No individual user data is accessible
Research use only — no commercial use
Attribution to MindFrame required in publications

Ready to start your research?

Email us with your research brief. We review all applications within 5 business days.