massive-api-client

massive-api-client is an async wrapper around parts of the official massive Python client.

It focuses on two things:

• async access to selected Massive.com REST APIs
• automatic sleeping and retrying when a rate limit is hit, which is especially helpful on the free tier

This package is still a work in progress, so it currently supports only a subset of endpoints. It reuses response models from massive.rest.models, which makes it easy to work alongside the official client.

Features

• Async client built on httpx.AsyncClient
• Reuses models from the official massive package
• Async pagination support through async for
• Built-in retry/sleep behavior for HTTP 429 Too Many Requests
• Small and focused API surface

Supported APIs

Reference

• list_tickers(...)
• get_ticker_details(ticker, ...)
• get_ticker_events(ticker, ...)
• list_ticker_news(...)
• get_ticker_types(...)
• get_related_companies(ticker)

Aggregates

• list_aggs(ticker, multiplier, timespan, from_, to, ...)
• get_grouped_daily_aggs(date, ...)

Indicators

• get_sma(ticker, ...)
• get_ema(ticker, ...)
• get_rsi(ticker, ...)
• get_macd(ticker, ...)

Installation

Install from PyPI:

pip install massive-api-client

Quick Start

import asyncio

from massive_api_client import MassiveAPIClient
from massive_api_client.config import MassiveClientConfig


async def main() -> None:
    client = MassiveAPIClient(
        MassiveClientConfig(
            massive_api_key="YOUR_MASSIVE_API_KEY",
        )
    )

    try:
        details = await client.get_ticker_details("AAPL")
        print(details.name)

        sma = await client.get_sma("AAPL", timespan="day", window=10, limit=5)
        print(sma.values[0].value)

        tickers = [
            ticker
            async for ticker in client.list_tickers(search="apple", limit=5, max_num_pages=1)
        ]
        print([ticker.ticker for ticker in tickers])
    finally:
        await client.close()


asyncio.run(main())

Pagination

The following methods return async iterators:

• list_tickers(...)
• list_ticker_news(...)
• list_aggs(...)

Example:

bars = [
    bar
    async for bar in client.list_aggs(
        ticker="AAPL",
        multiplier=1,
        timespan="day",
        from_="2024-01-01",
        to="2024-01-31",
        max_num_pages=1,
    )
]

max_num_pages is available as a defensive guard for paginated endpoints. It is default to 10000 pages.

Rate Limit Handling

When Massive returns HTTP 429, the client will:

1. sleep for rate_limit_sleep_secs
2. retry the same request
3. repeat until rate_limit_max_retries is reached

If the request is still rate-limited after all retries, a RateLimitException is raised.

Default settings:

• rate_limit_sleep_secs=60.0
• rate_limit_max_retries=3

Configuration

Use MassiveClientConfig to configure the client:

from massive_api_client.config import MassiveClientConfig


config = MassiveClientConfig(
    massive_api_key="YOUR_MASSIVE_API_KEY",
    api_base_url="https://api.massive.com",
    timeout_secs=30.0,
    rate_limit_sleep_secs=60.0,
    rate_limit_max_retries=3,
)

Fields:

• massive_api_key: required Massive API key
• api_base_url: defaults to https://api.massive.com
• timeout_secs: request timeout in seconds, default to 30 seconds
• rate_limit_sleep_secs: how long to wait before retrying after HTTP 429, default to 1 minute
• rate_limit_max_retries: maximum retry attempts after HTTP 429, default to 3 - meaning together with the initial call, it will call the API upto 4 times

Return Types

Responses are deserialized into models from the official massive package, for example:

• Ticker
• TickerDetails
• TickerNews
• TickerTypes
• Agg
• GroupedDailyAgg
• SMAIndicatorResults
• EMAIndicatorResults
• RSIIndicatorResults
• MACDIndicatorResults

This means you can keep using the official model types while switching to an async workflow.

Errors

• RateLimitException: raised when rate-limit retries are exhausted
• BadResponseException: raised when the API response is missing an expected field
• httpx exceptions: non-rate-limit HTTP failures are surfaced by httpx

Notes

• This project is intentionally small (focusing on author's personal requirements) and currently covers only part of the Massive API.
• The client is not an async context manager right now, so close it explicitly with await client.close().

Development

pip install -e . pytest
pytest

License

MIT