QuanterraOS
Check a market → Enter Cockpit
Developer Platform • OpenAPI 3.1 • MCP Server

Programmatic Telemetry & MCP for AI Agents

Equip your quantitative strategies, autonomous agents, and bots with true-cost fee calculations, live CME CF BRTI 60s TWAP progression, and empirical calibration scores across Kalshi and Polymarket.

Developer Quickstart
Query live prediction market friction, settlement benchmarks, and empirical calibration.
# 1. Fetch live settlement radar & CME CF BRTI 60s TWAP dispersion
curl -s "https://quanterraos.com/api/v1/radar/KXBTC15M" \
  -H "Authorization: Bearer YOUR_API_KEY"

# 2. Compute non-linear Kalshi taker fee & required breakeven win rate
curl -X POST "https://quanterraos.com/api/v1/fees/check" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{"venue":"kalshi","product":"KXBTC15M","price":0.50,"contracts":100}'
OpenAPI 3.1.0

OpenAPI 3.1 Specification

Complete machine-readable JSON schema defining all REST telemetry endpoints, request models, taker fee formulas, and response structures.

Model Context Protocol

Remote MCP Server

Connect Claude Desktop, Cursor, or sovereign AI agents directly to QuanterraOS tools via standardized JSON-RPC Model Context Protocol.

LLM Standard Context

llms.txt & agents.md

Standardized, lightweight Markdown files engineered for coding LLMs to ingest fee schedules, settlement logic, and system prompt constraints without HTML scraping.

Distribution Engine

Embeddable Widgets & SDK

Five pre-built iframe and JavaScript widgets including fee calculators, dispersion tickers, live countdowns, and cross-venue comparators with white-label flags.

AI Agent Configuration

Add QuanterraOS to Claude Desktop or Cursor

Add the remote server configuration below to your claude_desktop_config.json or .cursor/mcp.json:

{
  "mcpServers": {
    "quanterraos": {
      "url": "https://quanterraos.com/api/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_API_KEY"
      }
    }
  }
}
Endpoints: JSON-RPC over HTTP • Supports tools: calculate_true_cost, get_spot_basis, get_calibration_metrics, simulate_order_friction.
API Tiers & Rate Limits
Transparent quotas enforced via standard HTTP headers (X-RateLimit-Limit, X-RateLimit-Remaining).
Tier Price Monthly Request Limit Latency / Freshness MCP & Agent Access Support SLA
Cadet (Free) $0 1,000 req/mo Delayed (15m/20m) Standard manifest Community (<72h)
Builder API $49/mo 50,000 req/mo Sub-second Real-Time Full Remote MCP Server Priority (<24h)
Institutional / Data Custom ($1,500+/mo) Unmetered Direct L2 WebSocket & Ticks Dedicated MCP Concierge Dedicated Slack (<1h)
Instant Sandbox Key

Generate Free Developer Test Key (Cadet Tier)

Instantly generate a client-side test key with 1,000 requests/month quota to start prototyping in minutes.

Quota: 942 / 1,000 requests remaining this period • Resets in 20 days • No credit card required.
Research Datasets & Parquet Corpora
Download prospective settlement archives and query them directly with DuckDB, Pandas, or Polars.

CME CF BRTI vs Spot Dispersion (Tick Level)

Parquet • 482 MB
Q1 2026 – Present • S3 URI: s3://quanterra-datasets/tick/brti_dispersion_2026.parquet

Sub-second orderbook top-of-book quotes across Coinbase, Kraken, Bitstamp, and Gemini paired with 60-second sliding TWAP calculations.

-- DuckDB SQL Example:
SELECT window_time, exchange, dispersion_bps FROM read_parquet('s3://quanterra-datasets/tick/brti_dispersion_2026.parquet') WHERE dispersion_bps > 15 LIMIT 10;

1,316-Window Prospective Calibration Corpus

CSV & Parquet • 34 MB
1,316 Settled BTC15M Windows • S3 URI: s3://quanterra-datasets/corpus/settled_windows_1316.parquet

Complete canonical empirical ledger recording pre-close probabilities, actual resolutions, Brier score decomposition, and fee drag.

Download Direct CSV Archive ↗
-- DuckDB SQL Example:
SELECT window_id, market_mid_p, outcome, (market_mid_p - outcome)^2 AS brier FROM read_parquet('s3://quanterra-datasets/corpus/settled_windows_1316.parquet');

Kalshi BTC 15M Orderbook Depth Snapshots

Parquet • 1.2 GB
Rolling 90 Days • S3 URI: s3://quanterra-datasets/snapshots/kalshi_btc_l2_depth.parquet

1-minute snapshots of L2 orderbook bid/ask tiers and spread depth for KXBTC15M and KXBTCD event contracts.

-- DuckDB SQL Example:
SELECT timestamp, bid_cents, ask_cents, (ask_cents - bid_cents) AS spread FROM read_parquet('s3://quanterra-datasets/snapshots/kalshi_btc_l2_depth.parquet');
CIRCUIT BREAKER: LOCKED_RULE_B5 ($0.00 CAPITAL RISK) OPENAPI 3.1 • MCP 2026-01-01 SPEC COMPLIANT
Legal & Regulatory Disclosures: QuanterraOS developer APIs are independent computational and statistical tools. They do not accept deposits, route transactions, execute live orders, or provide investment advice. All API keys enforce permanent $0 live capital isolation. Kalshi, CME Group, and Polymarket are trademarks of their respective owners.