What replyCount tells you before you pay
Instagram Comments Scraper charges $0.0075 for the post and $0.0005 for each comment past the first fifteen. Replies are comments, so switching threads on multiplies the row count by something only the post knows. The replyCount on the comments you already have is the cheapest possible forecast of that.
Two prices, one of which you control
The billing has a shape worth internalising before you plan anything:
- $0.0075 per post, and the post itself is not a dataset item
- $0.0005 per comment, but only past the first fifteen
So a post with fifteen comments or fewer costs three quarters of a cent and nothing else. A post with two hundred costs $0.0075 plus 185 × $0.0005, which is a shade under ten cents.
The first number is fixed by how many posts you name. The second is fixed by the posts themselves, and that is the one that surprises people when threads come into it.
Replies are comments
There is no separate reply fee, which sounds like good news and is mostly neutral. A reply is a row, rows past fifteen cost $0.0005, and a post with a hundred top-level comments averaging three replies each is four hundred rows rather than a hundred.
The rate did not move. The run got four times bigger.
The estimate is already in your data
Every top-level comment carries a replyCount. Sum it and you have the exact
size of the threaded run, before you pay for it:
top = comments(post_url, include_replies=False)
replies_waiting = sum(c["replyCount"] for c in top)
billable = max(len(top) + replies_waiting - 15, 0)
print(f"{len(top)} comments, {replies_waiting} replies")
print(f"threaded run: ${0.0075 + billable * 0.0005:.4f}")The measurement run is not free, but at three quarters of a cent plus whatever the top level costs, it is cheap enough that guessing is the expensive option.
Where the fifteen matters most
The free allowance is per post, not per run. That changes which shape of job is good value:
- One post, 500 comments: $0.0075 + 485 × $0.0005 = $0.25
- 50 posts, 15 comments each: 50 × $0.0075 = $0.375, no item charges at all
Seven hundred and fifty rows across fifty conversations for thirty-seven cents, against five hundred rows from one conversation for twenty-five. If the question is “what is the reaction across this campaign” rather than “what happened under this one post”, the wide sweep is both cheaper per row and a better instrument.
A sequence that does not waste money
Pull the top level across the posts you care about. Sum replyCount to find
where the actual discussion is. Then re-run with threads on for the two or
three posts that carry it, rather than for all fifty.
Most campaign comment sections are shallow. The one that is not is usually the one worth reading, and finding it costs about a cent.
Questions people ask
How much does it cost to scrape Instagram comments?
$0.0075 for the post, which includes roughly the first fifteen comments, then $0.0005 for each comment beyond that. The post itself is not charged as a dataset item, so a post with fifteen comments or fewer costs three quarters of a cent flat.
Do replies cost extra?
Not at a different rate. A reply is a dataset item like any other comment, so turning threads on does not change the price per row. It changes how many rows there are, and the multiplier is a property of the post rather than anything you set.
How do I estimate a threaded run before paying for it?
Pull the top level first, then sum the replyCount field across those comments. That total is what the threaded run will add, and the top-level pull that produced the estimate is cheap enough to treat as a measurement.