MarketParquet vs Twelve Data
> multi-asset real-time API with 100+ technical indicators
Twelve Data is a clean, multi-asset API -- global stocks, forex, crypto, indices, 100+ built-in indicators, real-time WebSocket -- metered by API credits. MarketParquet is bulk US OHLCV as Parquet you download once. Different tools; here's how they line up for historical backtesting.
Side-by-side
| Twelve Data | MarketParquet | |
|---|---|---|
| Coverage | US + global stocks, ETFs, forex, crypto, indices, commodities | US stocks, ETFs, futures |
| Historical depth | daily to ~1980; intraday only from ~2022 | stocks/ETFs from 2000, futures from Dec 2007 |
| Format | JSON/CSV via REST + WebSocket | Apache Parquet, by-date partitioned |
| Free tier | 8 API credits/min, 800/day | recent daily sample, unlimited downloads |
| Real-time data | yes (delayed on the free tier) | no -- end-of-day batch only |
| Built-in indicators | 100+ (RSI, MACD, etc.) server-side | no -- compute with pandas/talib |
| Pricing | subscription: ~$29-$329/mo (API-credit model) | one-time $79 |
| Split / dividend adjusted | split-adjusted by default; dividends optional | yes |
| Includes delisted tickers | opt-in flag; completeness not guaranteed | yes -- included |
Where Twelve Data wins
- Broad multi-asset coverage through one clean, unified API (stocks, ETFs, forex, crypto, indices, commodities).
- Excellent developer experience -- strong docs, SDKs in many languages, and spreadsheet add-ins.
- Strong low-latency WebSocket streaming plus 100+ server-side technical indicators.
- Transparent, self-serve, no-contract pricing with a usable free tier for prototyping.
Where MarketParquet wins
- Bulk Parquet for backtesting vs an API-credit model -- no 8-calls/minute ceiling, no paging thousands of symbols.
- One file per trading day (all tickers) for cross-sectional research, cached locally.
- One-time $79 vs a recurring monthly subscription.
- Deeper intraday history: 1-minute US bars from 2000, vs Twelve Data's intraday starting around 2022.
Which to pick
Pick Twelve Data if you need global markets, real-time quotes, built-in technical indicators, or forex and crypto in one API.
Pick MarketParquet if you're backtesting US stocks, ETFs, or futures on bulk local Parquet and want deep 1-minute history without API-credit limits.
FAQ
Twelve Data vs MarketParquet for backtesting?
Twelve Data is a real-time, multi-asset API with built-in indicators, metered by API credits (8/minute on the free tier). MarketParquet is bulk Parquet you download once and query locally with no rate limits. For US equity/ETF/futures backtesting, MarketParquet is simpler and a one-time $79; Twelve Data wins if you need global coverage, real-time, or forex/crypto.
Does Twelve Data offer bulk downloads or Parquet?
No -- Twelve Data is an API (JSON/CSV over REST plus WebSocket), so historical data is metered per API credit and there's no bulk flat-file or Parquet export. MarketParquet ships Parquet partitioned by trading day for fast local loads.
How far back does Twelve Data's history go?
Daily data for major US stocks goes back to around 1980, but intraday (1-minute and similar) history only starts around 2022. MarketParquet provides 1-minute-to-daily US history from 2000 (futures from December 2007), so it's deeper for intraday backtesting.
Is Twelve Data free?
It has a free Basic tier limited to 8 API credits/minute and 800/day for non-commercial use, with some real-time data delayed. MarketParquet offers a free recent-daily sample and a one-time $79 complete archive you keep forever.
> Pricing and feature snapshots reflect public information at time of writing and may shift. Always confirm against Twelve Data's own pricing page before committing.