1. API
  2. Stock analytics
  3. Borrow rate

Stock analytics API, GET /borrow-rate/{ticker}

Is NVDA hard to borrow, and is the borrow rate spiking?

Borrow-rate level, class, percentile and the time series with spike detection. Returns the current 30-day and 2-year borrow rates, the class and percentile, a squeeze score, spike detection with severity, sector comparison and the daily series.

PremiumPlan and above, from $55/mo
1MCP tool that calls it
2026-09-08Sample captured from the live endpoint

The answer, in one call

Units first,
then the request.

Borrow rates in percent per year; percentiles 0 to 100.

Options-implied borrow from the data feed; spikes are z-scores against the trailing series.

curlBearer token from Account, then API Access
curl -H "Authorization: Bearer avmcp_YOUR_TOKEN" \
     "https://apexvol.com/api/mcp/data/borrow-rate/NVDA"
Pythonrequests, nothing else
import requests

headers = {"Authorization": "Bearer avmcp_YOUR_TOKEN"}
r = requests.get("https://apexvol.com/api/mcp/data/borrow-rate/NVDA", headers=headers)
r.raise_for_status()
data = r.json()["data"]
print(data["spike_detected"])
In Claude, Cursor or ChatGPTthrough the MCP server, no code
You

Is NVDA hard to borrow right now?

The assistant calls get_ticker_analytics and answers from the JSON below.

Tokens come with every paid or trial plan. Sign in and create one under Account, then API Access, or start a trial. Your ticker, your budget.

GET /api/mcp/data/borrow-rate/NVDA200, application/json, captured 2026-09-08
{
  "success": true,
  "data": {
    "current": {
      "avg_opt_volume": 3290793.0,
      "best_etf": "XLK",
      "borrow2yr": 3.5445,
      "borrow30": 3.9024,
      "borrow_class": "ELEVATED",
      "borrow_pctile": 98.8,
      "chg_1m": 5.05,
      "chg_1w": 5.86,
      "iv_percentile": 12.0,
      "market_cap": 5561054100000.0,
      "open_interest": 16025398.0,
      "options_volume": 5455875,
      "price": 230.3,
      "put_call_ratio": 0.5371,
      "residualRate": null,
      "sector": "Technology",
      "stock_loan_revenue": 898.72,
      "stock_volume": 6443150.0,
      "synthetic_short_cost": 8.68,
      "term_spread": 0.3579,
      "vrp": -11.87
    },
    "earnings_dates": [
      "2025-11-19",
      "2026-02-25",
      "2026-05-20"
    ],
    "next_earnings": null,
    "sector_comparison": {
      "sector_avg": 4.8474,
      "sector_count": 488,
      "sector_rank": 297,
      "vs_sector_ratio": 0.81
    },
    "spike_detected": false,
    "spike_ratio": 1.2836,
    "spike_severity": "MODERATE",
    "spike_z_score": 1.92,
    "squeeze_score": 26,
    "ticker": "NVDA",
    "timeseries": [
      {
        "borrow2yr": 2.8857,
        "borrow30": 3.3621,
        "borrow30_ma20": null,
        "date": "2025-09-05",
        "is_spike": false,
        "residualRate": null,
        "term_spread": 0.4764
      },
      {
        "borrow2yr": 2.7273,
        "borrow30": 3.1602,
        "borrow30_ma20": null,
        "date": "2025-09-08",
        "is_spike": false,
        "residualRate": null,
        "term_spread": 0.4329
      },
      {
        "borrow2yr": 2.655,
        "borrow30": 3.2406,
        "borrow30_ma20": null,
        "date": "2025-09-09",
        "is_spike": false,
        "residualRate": null,
        "term_spread": 0.5856
      }
    ],
    "velocity": {
      "borrow_chg_1m": 0.6,
      "borrow_chg_1w": 0.95
    }
  }
}

Every response is wrapped as {"success": true, "data": {...}}. Lists in the sample are cut to a few rows so it fits on a page; the shape is exactly what your code receives.

Parameters

Bounded,
and the bounds are stated.

Out-of-range numbers are clamped to the documented range, never silently changed to something else. A missing required field returns 400 with the reason.

NameInDefaultRangeMeaning
ticker requiredpathany covered symbolThe underlying, upper case. Use /search to check coverage.
daysquery25220 to 1000History length.

Field by field

What each
number means.

current.borrow30, borrow2yr, borrow_class, borrow_pctile
The rate now.
spike_detected, spike_severity, squeeze_score
The alert.
sector_comparison
Versus the sector.
timeseries[].date, borrow30, is_spike
The history.

Access, limits, errors

Before the
first call.

Plan
Premium Premium and above. Tokens are issued to paid and trial accounts from $55 a month; each endpoint follows the tier of its web feature. An endpoint above your plan answers 402 and names the plan it needs.
Auth
Bearer token, prefix avmcp_, created under Account, then API Access. Shown once, hashed at rest, rotates in one click. Always call https://apexvol.com, never www.
Limits
60 requests a minute and 1,000 an hour per token, plus a monthly allowance by plan. Every response carries the remaining counts in X-RateLimit-* headers; a 429 carries Retry-After in seconds.
Errors
401 no or revoked token; 429 Retry-After seconds; 403 Premium plan required; 424 data feed unavailable, retry. The full list with payload shapes is on the errors page.
Since
API 1.0. Units, bases and timestamps follow the conventions shared by every endpoint; changes are logged in the API changelog.
Machine-readable
This page as Markdown, the family as one file, the whole API as llms-full.txt or OpenAPI 3.1.

The same data, in plain English

Ask for it
in an MCP client.

Connect the ApexVol MCP server and the assistant calls get_ticker_analytics by name. One URL, a sign-in, no token to paste.

Prompts that hit this endpointcopy one into Claude, Cursor or ChatGPT
You

Is NVDA hard to borrow right now?

You

Which of my shorts has a borrow spike?

You

Show the borrow rate history for GME.

More in the prompt library. Set the server up in Claude Desktop, Claude Code, Cursor, ChatGPT, VS Code, Windsurf or Gemini CLI.

Included with
every paid plan.

From $55 a month, tier-matched: your token queries the data your plan includes and Pro unlocks the full surface.

Real market data, not a sandbox. See it live on AAPL.

7 days free, cancel anytime Card required · no charge for 7 days
Start trial →