Get Statistics Trending ​

GET https://openqiao.com/api/v1/management/statistics/trending

Rank models by a combined growth × magnitude score. Returns the top N models with the strongest upward momentum (including brand-new models that just spiked), each with its current/previous comparison values, growth rate, and a per-day detail series for sparkline rendering.

Use this endpoint to surface breakout models, build "Trending" leaderboards, or monitor momentum shifts across the catalog.

Ranking score: score = log1p(today) × (today + α) / (yesterday_eff + α), where today is the current-window volume, yesterday_eff is the previous-window volume (new models fall back to a dynamic baseline), and α is an additive smoothing constant. Multiplying the growth signal by a magnitude weight surfaces models that are both rising and sizable, while suppressing small-base noise.

Authentication ​

Authorization Header * ​

Authorization: Bearer <OPENQIAO_MANAGEMENT_API_KEY>

• Name: Authorization
• Format: Bearer <API_KEY>
• Description: A Management API Key created in the 开桥 Console

Rate Limiting ​

Each endpoint has its own independent rate limit counter. The maximum number of requests per minute is configured at the platform level. Exceeding the limit returns a 422 error.

Parameters ​

metric string * ​

Which metric to rank growth by.

• tokens — total input + output token count
• requests — total request count
• cost — list-price cost in USD

bucket_width string * ​

Comparison window for the growth rate.

• 1d — ending_at vs. the previous day; detail series covers the last 7 days
• 1w — last 7 days vs. the prior 7 days (both summed); detail series covers the last 14 days

ending_at string ​

Anchor date (inclusive), the end of the comparison window, in YYYY-MM-DD format.

• Default: yesterday (T-1, UTC)
• Must not be later than yesterday (T-1, UTC); data is aggregated daily and the current day becomes available the next day
• Must not be earlier than the data start date 2025-09-29

limit integer ​

Top N models to return.

• Default: 10
• Max: 50

Constraints ​

• Models whose current-window volume falls below a dynamic threshold are excluded, filtering out long-tail small-base noise.
• New models without a previous-window baseline fall back to a dynamic baseline, so they compete fairly on their current magnitude (neither pushed to the top for lack of a baseline, nor excluded outright).
• The detail window is fixed by bucket_width (1d → 7 days, 1w → 14 days) and cannot be customized.

Returns ​

data.metric string ​

Echo of the requested metric ("tokens", "requests", or "cost").

data.bucket_width string ​

Echo of the requested bucket width ("1d" or "1w").

data.ending_at string ​

Anchor date (YYYY-MM-DD).

data.entries array ​

Models sorted by combined score (growth × magnitude), descending. Each entry contains:

• rank integer — 1-based rank
• model string — Model slug (例如, qwen/qwen3-coder)
• label string — Model display name
• author string — Author/provider slug (例如, qwen)
• author_label string — Author display name
• icon_url string — Model (or author fallback) icon URL
• value number — Current-window volume (1d: anchor-day sum; 1w: last-7-day sum)
• total_value number — Total over the display window, identical to value. Use this for display.
• previous_value number — Effective baseline for the previous window: for existing models, the real previous-window volume (1d: previous day; 1w: prior 7 days); for new models, a dynamic fallback baseline (not a real historical value)
• growth_rate number — Additively smoothed period-over-period growth rate, rounded to 2 decimals. Because of smoothing it is not strictly equal to (value - previous_value) / previous_value; do not back-derive it
• growth_label string — Pre-formatted percentage label (例如, "+40615.61%")
• data_points array — Per-day detail series (ascending):
  • period string — Business day YYYYMMDD
  • value number — Metric value for that day

curl -G https://openqiao.com/api/v1/management/statistics/trending \
  -H "Authorization: Bearer $OPENQIAO_MANAGEMENT_API_KEY" \
  -d metric=cost \
  -d bucket_width=1d \
  -d ending_at=2026-06-21 \
  -d limit=10

{
  "success": true,
  "data": {
    "metric": "cost",
    "bucket_width": "1d",
    "ending_at": "2026-06-21",
    "entries": [
      {
        "rank": 1,
        "model": "qwen/qwen3-coder",
        "label": "Qwen3-Coder",
        "author": "qwen",
        "author_label": "Qwen",
        "icon_url": "https://cdn.marmot-cloud.com/storage/openqiao/2026/04/01/qeMamJm/Property-1Qwen.svg",
        "value": 1.63367713,
        "total_value": 1.63367713,
        "previous_value": 0.00401241,
        "growth_rate": 406.16,
        "growth_label": "+40615.61%",
        "data_points": [
          { "period": "20260615", "value": 0.00005835 },
          { "period": "20260616", "value": 0.26396599 },
          { "period": "20260617", "value": 1.44900805 },
          { "period": "20260618", "value": 13.43269683 },
          { "period": "20260619", "value": 16.53876505 },
          { "period": "20260620", "value": 16.87021709 },
          { "period": "20260621", "value": 1.63367713 }
        ]
      }
    ]
  }
}