TickDB: Unified Real-Time Market Data API for Stocks, Forex, and Crypto

One Connection for Six Market Types

TickDB positions itself as a single-endpoint replacement for managing multiple market data subscriptions. Without a unified API, a developer building a multi-asset dashboard would connect separately to a US equities feed, a Forex provider, a crypto exchange WebSocket, and a data source for Hong Kong and mainland China stocks. Each connection carries its own authentication, symbol format, and rate limit.

TickDB consolidates those into one API. The supported markets are: Forex and precious metals (using standard currency-pair symbols like GBPUSD and XAUUSD), US stocks on NYSE and NASDAQ (AAPL.US format), HK Exchange securities (700.HK format), Shanghai and Shenzhen A-shares (600519.SH format), global indices (SPX), and cryptocurrency pairs (BTCUSDT). The README states the coverage reaches 37,527 symbols across these six categories.

The access channels are REST API, WebSocket, CLI, MCP server, and a Skill for AI coding assistants. The same data flows through all five channels; the choice of access method depends on the application. REST suits batch requests and historical data retrieval. WebSocket suits real-time dashboard updates. MCP and Skill integrations let AI agents query live prices through natural language.

REST API Endpoints and Symbol Format

The REST API base URL is https://api.tickdb.ai. All requests require an API key in the X-API-Key header:

curl -H "X-API-Key: YOUR_API_KEY" \
     "https://api.tickdb.ai/v1/market/ticker?symbols=AAPL.US,700.HK,BTCUSDT"

Four main endpoints cover the primary data types. The /v1/market/kline endpoint retrieves historical candlestick data with symbol, interval, and limit parameters. The /v1/market/ticker endpoint returns real-time price data for one or more symbols in a single request. The /v1/market/depth endpoint returns order book data. The /v1/market/trades endpoint returns recent trade history. A fifth endpoint, /v1/symbols/available, returns the list of tradable instruments for a given market:

curl -H "X-API-Key: YOUR_API_KEY" \
     "https://api.tickdb.ai/v1/symbols/available?market=HK&limit=10"

The symbol format encodes the market in the suffix: .US for US stocks, .HK for Hong Kong stocks, .SH or .SZ (implied) for A-shares, no suffix for Forex pairs and crypto.

WebSocket Subscriptions and Latency

The WebSocket endpoint is wss://api.tickdb.ai/v1/realtime, authenticated via an api_key query parameter. Three channel types are supported: ticker for continuous price updates, depth for order book changes, and trade for real-time execution records. The README gives this subscription example:

const ws = new WebSocket('wss://api.tickdb.ai/v1/realtime?api_key=YOUR_API_KEY');

ws.onopen = () => {
    ws.send(JSON.stringify({
        cmd: 'subscribe',
        data: { channel: 'ticker', symbols: ['BTCUSDT'] }
    }));

    ws.send(JSON.stringify({
        cmd: 'subscribe',
        data: { channel: 'depth', symbols: ['BTCUSDT'] }
    }));

    ws.send(JSON.stringify({
        cmd: 'subscribe',
        data: { channel: 'trade', symbols: ['BTCUSDT'] }
    }));
};

The README states end-to-end latency for WebSocket data as approximately 10 to 50 milliseconds. This figure comes from the README and is not independently verified here.

MCP Server and CLI for AI Agent Integration

The repository includes a Python MCP server under mcp/ with 13 market data tools. The hosted endpoint at mcp.tickdb.ai runs the same code published in that directory. To connect Claude Code, Cursor, Kiro, or other MCP-compatible clients to the hosted server, add a JSON entry to the MCP client configuration:

{
  "mcpServers": {
    "tickdb": {
      "type": "http",
      "url": "https://mcp.tickdb.ai/",
      "headers": {
        "X-TickDB-Key": "YOUR_API_KEY"
      }
    }
  }
}

The MCP server source requires Python 3.11 or later, ships with a Dockerfile, and includes 46 unit tests. CI workflow runs against those tests. Teams that need to self-deploy the MCP server can use the mcp/ directory directly.

The CLI provides a terminal interface for quick queries and scripting. Install globally and configure the API key:

npm install -g tickdb
tickdb config set-key YOUR_API_KEY
tickdb ticker BTCUSDT,XAUUSD

A Skill integration for AI coding assistants is also available. Installing the skill via npx clawhub@latest install tickdb-market-data provides a trial key automatically, covering 72 popular symbols without requiring registration.

Limitations and What the Open-Source Repository Provides

The open-source repository contains the MCP server source, the Skill configuration files in SKILL/, and documentation. It does not include the data infrastructure itself. TickDB is a hosted API service; the data endpoints at api.tickdb.ai and the WebSocket server are not part of the open-source release.

The README states A-share data covers Shanghai and Shenzhen exchanges but does not document the specific data providers, update frequency outside of the stated WebSocket latency, or the historical depth available per symbol. It also does not describe SLA guarantees or rate limits, pointing instead to the full documentation at docs.tickdb.ai.

The last push to the repository was on 2026-06-20, when the MCP server reached version mcp/v0.1.3. The project is not archived. Community support runs through GitHub Issues, a Telegram channel, and email at [email protected].

How TickDB Compares to Polygon.io and Similar Feeds

Polygon.io is a well-known US market data provider offering REST and WebSocket APIs for equities and options, with a free tier for delayed data and paid tiers for real-time. Its coverage is primarily US markets; multi-market access requires additional subscriptions or a different provider for each region.

TickDB's differentiation is the single-endpoint multi-market scope: one API key, one authentication header, and one WebSocket connection covering US stocks, HK stocks, A-shares, Forex, metals, indices, and crypto. For teams building global or multi-asset applications, that removes the integration work of connecting to separate regional providers.

The trade-off is that TickDB is a newer service, and the open-source portion is limited to the MCP server and Skill wrappers. Teams with requirements for exchange-certified data, tick-level compliance data, or SLA commitments that need to be contractual should evaluate the full commercial terms at tickdb.ai and compare those against established providers for their specific markets.

Editorial conclusion

TickDB suits developers who need real-time market data across multiple asset classes from a single API and want MCP or CLI integration for AI agent workflows. The last push was on 2026-06-20, and the MCP server reached version 0.1.3 on the same date. The open-source MCP code in the mcp/ directory is available under MIT, so teams that cannot use the hosted endpoint at mcp.tickdb.ai can self-deploy from the published source.

Frequently asked questions

Official sources

1. License: MIT
2. Project website
3. README
4. Releases
5. TickDB/tickdb-unified-realtime-marketdata-api on GitHub