Google Finance Scraper — Google Finance API for live stock quotes

A google finance scraper that returns the live quote and the whole key-statistics block Google Finance shows beside it. One dataset row per ticker: price, absolute and percentage change, direction, open, day high and low, market cap, average volume, volume, dividend and yield, quarterly dividend, ex-dividend date, P/E, the 52-week range, EPS, beta, shares outstanding and headcount. Send a watchlist, get JSON you can sort, chart or pipe into a report.

Pass AAPL or AAPL:NASDAQ. A ticker that names its exchange keeps it; one that does not falls back to Default exchange, which starts at NASDAQ. For anything outside US tech, name the market: ASML:AMS, SHEL:LON, SAP:ETR, 7203:TYO. ASML on NASDAQ and ASML on AMS are different listings at different prices in different currencies.

What this Google Finance scraper returns per quote

Every large figure comes back twice. market_cap is 4910000000000 and market_cap_text is "4.91T". The same holds for volume, average volume, shares outstanding and headcount. A market cap of "4.91T" is unsortable in a spreadsheet; a market cap of 4910000000000 is unreadable in a report. You get both and pick per column.

currency is read off the page rather than inferred from the exchange, because a listing can quote in a currency the exchange's country does not use. London names in pence come back as GBX with the figures Google shows.

as_of carries the timestamp exactly as the page states it, including the market's own timezone (Oct 7, 4:00:01 PM GMT-4). Outside trading hours that is the closing price; the after-hours figure is a different number on the same page and is not reported as the price.

fetched_at is the ISO 8601 UTC moment this row was pulled.

Where a cell is empty

Google writes - where it has no figure, and those arrive as null rather than zero. A P/E of null means the company has no meaningful one, usually because it has no earnings; a P/E of 0 would be a claim nobody made. previous_close and cdp_score appear on some listings and not others.

Input

field	type	default	notes
tickers	array	required	AAPL or AAPL:NASDAQ, one per entry
exchange	string	NASDAQ	Fallback market for bare tickers
language	string	en	Two-letter interface language
country	string	us	Two-letter market code for the request
concurrency	integer	3	Parallel quotes, 1–10
attempts	integer	8	Tries per ticker with a fresh session, 1–30

Output

One row per input ticker, with found, attempts, duration_seconds, error, position and url on every row. summary repeats the run counts (tickers, quoted, withMarketCap, withPeRatio, defaultExchange, sessionsInPool, proxyConfigured, proxyPinnable, fetchedInSeconds, errorCount). errors lists not_parsed and blocked per ticker.

A real run, {"tickers": ["AAPL:NASDAQ", "MSFT:NASDAQ"]} with everything else at its default, pulled 8 October 2026 — the first row and the summary of that two-row run, as examples/prefill.json records them:

{
  "items": [
    {
      "symbol": "AAPL",
      "exchange": "NASDAQ",
      "name": "Apple Inc",
      "currency_symbol": "$",
      "currency": "USD",
      "price": 336.67,
      "change": 3.04,
      "change_percent": 0.91,
      "direction": "up",
      "as_of": "Oct 7, 4:00:01 PM GMT-4",
      "open": 336.96,
      "day_high": 338.67,
      "day_low": 332.78,
      "previous_close": null,
      "market_cap": 4910000000000,
      "average_volume": 40200000,
      "volume": 34150000,
      "dividend_yield": 0.32,
      "quarterly_dividend": 0.27,
      "ex_dividend_date": "Aug 10, 2026",
      "pe_ratio": 38.59,
      "week52_high": 345.34,
      "week52_low": 243.42,
      "eps": 8.72,
      "beta": 1.07,
      "shares_outstanding": 14590000000,
      "employees": 166000,
      "cdp_score": null,
      "url": "https://www.google.com/finance/quote/AAPL:NASDAQ?hl=en&gl=us",
      "found": true,
      "attempts": 1,
      "duration_seconds": 2.57,
      "error": null,
      "position": 1,
      "fetched_at": "2026-10-08T03:15:11.823Z",
      "market_cap_text": "4.91T",
      "average_volume_text": "40.20M",
      "volume_text": "34.15M",
      "shares_outstanding_text": "14.59B",
      "employees_text": "166K"
    }
  ],
  "errors": [],
  "summary": {
    "tickers": 2,
    "quoted": 2,
    "withMarketCap": 2,
    "withPeRatio": 2,
    "defaultExchange": "NASDAQ",
    "language": "en",
    "sessionsInPool": 60,
    "proxyConfigured": true,
    "proxyPinnable": true,
    "fetchedInSeconds": 4.9,
    "errorCount": 0
  }
}

The second row of the same run is MSFT at 529.76, pe_ratio 29.52, week52_low 349.2, duration_seconds 1.08. Prices and volumes move every minute: the row above is what the page said at that moment, not a figure the Actor rounds or caches.

Errors

Code	Meaning
blocked	No attempt got a quote page back
not_parsed	A quote page arrived with no price on it
bad_input	No ticker was given
crash	Unexpected error inside the run; the run itself fails

blocked and not_parsed are different problems. The first is worth retrying (raise Attempts per ticker); the second means the page came back but the headline moved. crash is not a parse problem: something broke that the actor did not expect.

Use cases for the Google Finance scraper

Watch a portfolio

Put every holding in tickers, run once a day, keep price, change, change_percent, direction and as_of. A row per position, in the order you sent them (position), so a diff against yesterday's run is a one-line join on symbol.

Screen a sector

Send a whole list — every member of an index, every supplier you track — and sort on pe_ratio, week52_low / week52_high, dividend_yield or change_percent. The numeric fields are numbers, so a spreadsheet or a SQL ORDER BY works without a cleanup pass first.

Feed a dashboard

market_cap, volume, eps and beta arrive typed, and the *_text variants arrive as Google displayed them, so a dashboard can show 4.91T next to a chart axis that needs 4910000000000. fetched_at is the freshness stamp to print in the corner of the panel.

Enrich a table you already have

You have ISINs, or a list of suppliers, or tickers out of a filing. Resolve them to symbol + exchange yourself, then use the run to attach price, name, currency and the key-statistics block as extra columns.

How this Google Finance scraper compares

Search for a Google Finance scraper and you get three kinds of answer: the page itself, a script you run yourself, and a hosted Google Finance API — SerpApi, SearchAPI and ScrapeBadger all publish one. This Actor sits in a fourth spot: hosted, but priced per event instead of per credit pack.

	Google Finance in a browser	Open-source script	Hosted Google Finance API	This Google Finance scraper
Returns rows you can store	No, it is a page	Yes, once it runs	Yes	Yes
Works without writing code	Yes	No	Yes	Yes, via the run form
Separate API key and credit plan	No	No	Yes	No, pay per event
Key statistics block beside the price	Yes	Depends on the script	Varies by vendor	Yes, every field, as number and as text
Many tickers in one request	No	Depends on the script	Varies by vendor	Yes, one row per ticker
A ticker that fails still returns a row	n/a	Usually stops	Usually an error	Yes, found: false with a code

The hosted APIs win on exotic coverage: other pages, other exchanges, longer history. This Actor is deliberately narrow — the Google Finance quote page and its key-statistics block, for a list of tickers, at $0.002 a quoted ticker.

FAQ

Is this a Google Finance API replacement?

It is a scraper and an API for the same public quote page: you send tickers, you get JSON rows with price and key statistics. Use it where you would otherwise poll a paid stock price API and still want the Google Finance figures and the displayed text beside them.

Which exchanges are supported?

Any listing Google Finance serves, written as SYMBOL:EXCHANGE — AAPL:NASDAQ, ASML:AMS, SHEL:LON, SAP:ETR. Bare tickers use your default exchange. The currency is taken from the page, so a LSE name in pence returns GBX, not an assumed GBP.

How fresh is the quote?

The row carries as_of exactly as the page states it and fetched_at in UTC. During market hours that is the live print; outside them it is the official close. After-hours prints on the same page are not reported as the price.

What happens when a ticker cannot be read?

You still get one row per ticker with found: false and a clear error or a not_parsed note in errors. Only quoted tickers are billed, so a bad symbol or a blocked attempt does not cost a quote event.

Can I use this as a share price API for a watchlist?

Yes. Send the whole list in tickers, set exchange once if it is one market, and read price, change, change_percent, direction and as_of per row. Numbers are sortable; *_text keeps the form Google displayed.

Pricing

Pay-per-event, the same model as before:

event	price	when
run_start	$0.002	once at the start of every run
quote	$0.002	per ticker that was actually quoted (found: true)

Empty results are free: a ticker with no price is not a quote event. The start fee still stands. Free-plan runs deliver at most 10 rows per run and are only charged for what you receive; upgrade to remove the cap or split the work across runs.