Early access is open. Free sample datasets are live now. Download a sample

Compare

Polygon.io (Massive) Alternative

Polygon.io is now called Massive, and polygon.io redirects to massive.com. The API is still one of the cleanest around. People who look for an alternative usually have a specific gripe: they need more than one asset class and don't want to pay per class, or they want ten years of minute bars as a file without keeping a subscription running. Here's an honest side-by-side.

Details

At a glance

Our column is planned. We're in early access, and only the free sample downloads are live.

Massive (formerly Polygon.io) marketdata.co (planned, early access)
Asset classes US stocks (incl. OTC), US options, indices, futures (CME, CBOT, NYMEX, COMEX), forex, crypto [1] US stocks and ETFs, options, forex, crypto, futures, indices, economic data
Delivery REST, WebSocket, flat files over S3 (gzipped CSV) [1] REST, WebSocket, CSV and Parquet dataset files
Pricing model Per asset class: individual plans $29 to $199 a month each; business plans $999 to $2,500 a month per class [2] One plan covers all asset classes: Free, Starter $29, Pro $79, Business $249 a month; datasets priced per file
History Stocks since 2003-09-10, forex since 2009, crypto since 2013, futures since 2017 [1] Varies by product and plan; published per product page
File downloads Flat files bundled with paid plans; not sold separately [1] One-off datasets you keep, CSV or Parquet
Free tier Basic: 5 calls a minute, end of day, 2 years [2] Free plan planned; sample files downloadable now

Sources: [1] https://massive.com/docs [2] https://massive.com/pricing

Details

Where they are strong

Massive does a lot right. One URL pattern serves bars for stocks, options, indices, forex and crypto. Every docs page states the plan, recency and history depth, and each one has a Markdown version and an llms.txt index, which matters when your coding assistant is reading the docs. Stock history back to 2003 is long, the S3 flat files are a sensible way to deliver bulk data, and they're adding options analytics like Greeks history and unusual trades. If you trade one asset class and want a mature API, they're a strong pick.

Side by side

Where we differ

One plan, every asset class. Massive prices stocks, options, futures and currencies separately, so stocks plus options in real time is two subscriptions. Our planned plans cover all asset classes at one price per tier.

Datasets without a subscription. Massive's flat files come with a paid plan and follow its history limit. We plan to sell a dataset once, for example one symbol's 1-minute history, and you keep the file after you cancel.

Parquet as well as CSV. Massive's flat files are gzipped CSV. Ours are planned in both CSV and Parquet, which is smaller and much faster for minute and tick data.

Long field names. Massive uses short keys (o, h, vw, n). We spell them out (open, high, vwap, trade_count), so a dataframe is readable without a lookup.

Samples you can download now from our sample data page.

Watch out

Migration notes

Field mapping from Massive's aggregate bars and stock trades flat files to our schemas:

Massive field marketdata.co field Note
ticker symbol On every row
results[].o / h / l / c open / high / low / close
results[].v volume
results[].vw vwap
results[].n trade_count
results[].t (Unix ms) timestamp Ours is ISO 8601 UTC
adjusted adjusted (intraday bars) Daily bars carry adjusted and unadjusted_* columns side by side
next_url next_url Same pagination idea
sip_timestamp (trades) timestamp Nanosecond precision kept
participant_timestamp participant_timestamp
id (trades) trade_id
price, size, exchange, conditions, tape, sequence_number, correction, trf_id same names Our tick schema

Example: the bar loader change.

import pandas as pd

def load_massive(body):
    df = pd.DataFrame(body["results"]).rename(columns={
        "o": "open", "h": "high", "l": "low", "c": "close",
        "v": "volume", "vw": "vwap", "n": "trade_count"})
    df["timestamp"] = pd.to_datetime(df["t"], unit="ms", utc=True)
    return df.drop(columns="t")

def load_marketdata(body):  # names already match
    df = pd.DataFrame(body["results"])
    df["timestamp"] = pd.to_datetime(df["timestamp"])
    return df

FAQ

Frequently asked questions

Is Polygon.io the same company as Massive?

Yes. Polygon.io rebranded as Massive, and polygon.io now redirects to massive.com. The API and data products carried over under the new name.

Is marketdata.co cheaper than Polygon.io?

For one asset class, our planned prices are similar: both start at $29 and both have a $79 tier. The difference shows up when you need several asset classes, since our plans cover all of them at one price.

Can I buy Polygon-style flat files without a subscription?

That's what our datasets are for. We plan to sell individual files, such as one symbol's minute bars or one day of trades, as one-off purchases in CSV or Parquet.

Is marketdata.co live?

Not yet. The free sample downloads work today; the API opens product by product during early access.

Details

Try it

Load a real sample file into the same code you use for Massive and see how much renaming disappears.