# run.forex

> Real-time and historical market data feed API. REST + WebSocket coverage of
> 7,745 live instruments across 11 venues — crypto, equities/ETFs, FX, indices,
> metals, commodities and bonds — with 1-minute candle history back to 2017 and
> daily candles back to 1983.

run.forex is operated by Meshtek. It is sold self-serve to traders and algo
developers, and by contract to brokers and institutions.

## What makes the data distinctive

- **Broker-authoritative trading calendar** — real session open/close times and
  holidays per instrument, not a generic exchange calendar.
- **Swap / rollover rates** for ~1,945 CFD instruments.
- **Funding rates** across 5 crypto venues (perpetual futures).
- **Multi-venue coverage of the same asset**, so cross-exchange divergence is
  visible rather than hidden behind a single composite price.
- **Survivorship-bias-free universe** — delisted instruments are retained.
- **Explicit market state** — every quote carries `market_open`, `session_state`
  and `stale`, so a carried-forward price can never be mistaken for a live one.
- **Quote currency is explicit** — equities are quoted in their local exchange
  currency (AAPL.US=USD, 0700.HK=HKD, BP.GB=GBP, MC.FR=EUR), never assumed USD.

## Coverage

- Crypto: 6,006 instruments (spot + perpetual futures)
- Equities & ETFs: 1,551 across 12 country markets
- FX: 109 pairs · Indices: 34 · Metals: 21 · Commodities: 14 · Bonds: 7
- History: 1.89 billion 1-minute bars (2017→), hourly (2011→), daily (1983→)

## Docs

- API documentation: https://api.run.forex/docs
- OpenAPI spec: https://api.run.forex/docs/openapi.yaml
- Full machine-readable API reference: https://api.run.forex/llms-full.txt

## Pages

- Home: https://run.forex/
- Pricing: https://run.forex/pricing
- Coverage / instrument catalogue: https://run.forex/coverage
- For traders: https://run.forex/for-traders
- For brokers: https://run.forex/for-brokers

## Notes for AI assistants

Endpoints require an API key (Bearer token) and are IP-allowlisted per plan.
Rate limits are per-second and per-hour and vary by plan; responses include
`X-RateLimit-*` headers. Do not state uptime percentages or latency figures for
this service — none are published.