Yelp Scraper – Businesses, Reviews & Photos
Pricing
from $5.00 / 1,000 business records
Yelp Scraper – Businesses, Reviews & Photos
Scrape Yelp businesses: full records with phone, website, hours, health scores and Yelp's own recent-demand counter, plus every photo and the filtered reviews Yelp sells through no API tier.
Pricing
from $5.00 / 1,000 business records
Rating
0.0
(0)
Developer
Simple Actors
Maintained by CommunityActor stats
0
Bookmarked
2
Total users
1
Monthly active users
a day ago
Last modified
Categories
Share
Yelp businesses as data. Give it a place to search, or paste business URLs, and get one row per business: the full record, the photo gallery, and the reviews Yelp's own filter took off the page.
It is built around one fact. Yelp ended free API access, and what it sells now starts at $229 a month for a tier that returns no photos and no review text at all. The middle tier is $299 and gives three photos and three truncated review excerpts. Even the $643-a-month top tier stops at twelve photos and seven excerpts. Several of the fields below — the recent-demand counter, the health scores, the outbound website URL, the filtered reviews — are sold at no tier, at any price.
What you get
Identity and location — name, Yelp id, alias, Yelp URL, alternate names, whether the listing is claimed, whether it is closed or closed until a date, street address, city, state, postcode, country, neighbourhoods, timezone, latitude and longitude.
Contact — phone, the business's own website (unwrapped from Yelp's redirect, so you get the real domain), menu URL, whether Yelp messaging is on and how fast the owner answers.
Classification and standing — categories, price band, Yelp's one-line summary, the owner's own "specialties" text, year established, star rating, review count, whether they advertise on Yelp, whether Yelp has a consumer alert on them, and when their details were last verified.
Demand — how many people looked this business up recently, and Yelp's own count of the searches it appeared in. This is the closest thing on Yelp to a live busyness signal, and it exists in no API tier.
Opening hours — one row per day per interval, including the days they are shut, with the timezone needed to read them.
Health inspections — score, grade, inspection date, the health department behind it, and any itemised violations. US and Canada.
Reviews Yelp filtered out (optional) — two separate groups, tagged
not_recommended and removed, and they give you different things. The
demoted ones come with their star rating, date and full text — this is
the corpus Yelp sells nowhere. The removed ones, deleted for breaking Yelp's
terms, come with their rating and date but no text: Yelp replaces the body
with a notice, so text is empty on those by design rather than by omission.
Neither group counts towards the public rating or review count. Every row also carries both totals, so you can see the gap
between the public review count and the real one — for the restaurant used in
testing, 6,137 public reviews sat alongside 1,040 demoted and 419 deleted.
Review highlights and amenities — Yelp's extracted recurring phrases with the number of reviews behind each, and the full "Amenities and more" panel.
Ask the Community — questions and their answers, with the total.
Photos (optional) — every photo's id, caption and dimensions, plus each resolution Yelp serves it at, so you can pick your own size.
What it does not do
It does not return Yelp's recommended reviews. Yelp loads those in the browser from an endpoint that needs a logged-in session, and no amount of page reading gets at them. The filtered reviews above are a different corpus — the one Yelp hides — and they are the reviews this actor returns. If you need the main review feed, this is not the tool.
Two ways to ask
Search a place. Set Where to something you would type into Yelp —
San Francisco, CABrooklyn, NY, London — and optionally What (pizza,
plumber). Each result is then opened as a full business page. Turn on Search
results only to skip that and get just the thin result-page row, which is much
cheaper and much faster.
Name the businesses. Paste Yelp URLs into Yelp business URLs, or just the
alias from the end of one. yelp.co.uk and yelp.ca addresses work too.
Both inputs are read together, so you can do either or both.
What it costs
You are charged for what you actually receive, not for the work behind it:
| You get | Price each | Per 1,000 |
|---|---|---|
| A search-result row | $0.002 | $2 |
| A full business record — all 64 fields | $0.005 | $5 |
| A filtered or removed review | $0.001 | $1 |
| A photo | $0.0005 | $0.50 |
So a thousand businesses as a lead list is $2, a thousand full records is $5, and a thousand records with forty filtered reviews each is $45. For comparison, Yelp's own API charges $229 a month before a single call and returns no review text at all; its $643-a-month tier returns seven truncated excerpts.
A row that could not be read is never charged. Nor is anything you switched off — reviews and photos are opt-in, and the two limit settings are your ceiling on each.
Reading the output
Every business is one row. A business that could not be read is also a row,
carrying error and errorDescription and no name field at all — so a
failed lookup can never be mistaken for a business with no name. Error codes are
bad_input (not a Yelp business address), not_found (no such business),
blocked and read_failed.
The run stays green either way. A bad URL or a missing business is an answer, not a failure, so it does not turn the run red — the error rows are the failure channel, and the run's status message counts them. If you schedule this, watch the rows rather than the run status.
Privacy
Reviewer names and their friend, review and photo counts are off by default. Reviewers are private individuals rather than businesses, and the review itself — rating, date, text — is what most callers need. Turn on Include who wrote each review only if you have a reason to hold personal data.
Unofficial
Not affiliated with, endorsed by, or connected to Yelp. Yelp's terms prohibit automated access to its site; you are responsible for your own use of this actor and for the data it returns.