Three public surfaces — historical request/response (MDDS gRPC), real-time streaming (FPSS), and whole-universe daily blobs (FLATFILES) — plus a local Greeks calculator, exposed in Rust, Python, TypeScript, and C++ from a single Rust core.
One Rust core, four bindings: Rust, Python (PyO3, abi3), TypeScript/Node.js (napi-rs), C++ (RAII header-only). Same API shape, typed results in each language's idiom. The C ABI in `ffi/` is also the supported integration path for any third-party Go/C consumer that wants to roll their own wrapper.
FPSS client with SPKI certificate pinning, SPSC ring buffer (131,072 slots), configurable reconnect policy. Subscribes to quotes, trades, open interest, and full-stream firehoses.
23 Black-Scholes Greeks plus an IV solver, computed in Rust with no server round-trip. The server-computed Greeks endpoints are also exposed.
Standalone CLI for one-off queries, an MCP server that exposes every generated historical endpoint plus three offline tools to any MCP-compatible client, and a REST + WebSocket server on port 25503.
cargo add thetadatadx tokio --features tokio/rt-multi-thread,tokio/macrospip install thetadatadx
# With DataFrame adapters
pip install thetadatadx[pandas]npm install thetadatadx# Build the FFI library once, then include sdks/cpp/include/thetadx.hppuse thetadatadx::{ThetaDataDxClient, Credentials, DirectConfig};
#[tokio::main]
async fn main() -> Result<(), thetadatadx::Error> {
let creds = Credentials::from_file("creds.txt")?;
let client = ThetaDataDxClient::connect(&creds, DirectConfig::production()).await?;
let quotes = client.stock_history_quote("AAPL", "20250115", "60000").await?;
for q in "es {
println!("{}: bid={} ask={}", q.date, q.bid, q.ask);
}
Ok(())
}from thetadatadx import ThetaDataDxClient, Credentials, Config
creds = Credentials.from_file("creds.txt")
client = ThetaDataDxClient(creds, Config.production())
quotes = client.stock_history_quote("AAPL", "20250115", "60000")
for q in quotes:
print(f"{q.date}: bid={q.bid:.2f} ask={q.ask:.2f}")import { ThetaDataDxClient } from 'thetadatadx';
const client = await ThetaDataDxClient.connectFromFile('creds.txt');
const quotes = client.stockHistoryQuote('AAPL', '20250115', '60000');
for (const q of quotes) {
console.log(`${q.date}: bid=${q.bid} ask=${q.ask}`);
}#include "thetadx.hpp"
auto creds = tdx::Credentials::from_file("creds.txt");
auto client = tdx::Client::connect(creds, tdx::Config::production());
auto quotes = client.stock_history_quote("AAPL", "20250115", "60000");
for (const auto& q : quotes) {
std::cout << q.date << ": bid=" << q.bid << " ask=" << q.ask << std::endl;
}| Axis | ThetaDataDx |
|---|---|
| Languages | Rust, Python, TypeScript, C++ |
| Historical endpoints | Full typed historical surface (plus 4 _stream SDK-only variants) |
| Real-time streaming | FPSS with SPKI pinning, SPSC ring, reconnect policy |
| Local Greeks calculator | 23 Greeks + IV solver in Rust |
| Async Python surface | *_async variant of every endpoint |
| DataFrame output | Arrow / polars / pandas via explicit conversion |
Historical decode runs in Rust and materialises typed structs (or an Arrow RecordBatch, zero-copy at the PyO3 boundary).