Instagram data.
Without the login.

Here it means one HTTP endpoint that returns public Instagram data as clean JSON: an account’s profile and the story highlights on it, the posts it has published, its reels with play counts, one post or reel in full, the public page of comments, Instagram’s own reel feed and the terms it is trending. You send a handle, a post URL or nothing at all with a Bearer token, and get back typed records. It is not Instagram’s own Graph API — there is no app review, no Business account requirement and no OAuth. Everything returned is what Instagram shows a visitor who is not signed in.

No. You send your FetchLayer key and nothing else. There is no Instagram login, no app registration and no approval step, because this Instagram scraper API reads the public site rather than a partner API. FetchLayer is not affiliated with, endorsed by, or sponsored by Instagram or Meta, and the Instagram name is used only to identify the platform the data comes from.

Twelve. Instagram serves a grid 12 posts at a time whatever limit says, so limit: 60 is read as five pages and costs five credits — and the response states exactly what it did as pagesFetched, so a bill is never a surprise. Up to 240 items and 20 pages in one request. This is Instagram’s page size, not ours, and pretending otherwise would just mean a silently larger bill.

About thirteen — the first page, and only the first page. Instagram publishes that much to a signed-out reader and no more, while still reporting a total in the thousands and setting hasMore true. Rather than let you loop on that, every response says it outright: “These are the 13 comments of the 4,334 the source reports that are public for this post. The source does not publish the rest to a signed-out reader, so there is no further page to read.” Paging again returns the same thirteen and bills again, so do not. Reply text is not public either; each comment reports only how many replies it has.

Because Instagram does not publish one to a reader who is not signed in. Follower and following counts come back normally; stats.postCount is null for every account, and the response carries a note saying to read the account’s posts and count them. It is stated rather than quietly zeroed, because a zero would be a wrong answer and a null with an explanation is a true one.

Two things, both free with a call you were already making. Story highlights: the profile response lists the id, title and cover of every highlight the account keeps, which is how an account organises itself over time — their contents are not public, and nothing here pretends otherwise. And play counts on reels: the reels listing is the only Instagram listing that carries engagement at all, so every reel comes back with playCount, likeCount and commentCount without a detail call each.

No, and it is worth being plain about why rather than listing it as “coming soon”. Instagram redirects a signed-out reader to a login page for hashtag pages, location pages and keyword search alike. Reading them needs authenticated Instagram accounts, which is a different product with a different risk profile — not a feature we left out. What Instagram does publish is its own reel feed and the terms it is promoting, and both are available here. A hashtag or place link still resolves through /resolve-url, which tells you what it names and that it cannot be read.

No. Instagram publishes none to a signed-out reader. What you do get is caption — the author’s own written post body — and altText, which is Instagram’s machine-written image description (“Photo by NASA on August 19, 2026. May be an image of text.”). Neither is a transcript of spoken audio, and the docs say so in the same words, because a caller who assumes otherwise builds something that quietly does the wrong thing.

Instagram does not publish them on a grid. A posts listing carries the caption, hashtags, mentions, alt text and every rendition, but no like count, no comment count and no posting date — those come from a detail call. Every record says which shape produced it in fieldsFrom, so a null is never mistaken for a zero, and the response says so in notes rather than leaving you to discover it. Reels are the exception: that listing does carry engagement.

One credit per page of results read, and every response states it as pagesFetched. A single-page request costs 1 credit whatever the page holds. Because Instagram’s page is twelve items, asking for sixty costs five. Parsing a URL with /resolve-url reaches Instagram not at all, reports pagesFetched 0 and is free. Instagram is in Beta while coverage expands, billed at half credit.

Every route is monitored 24/7 and a failure alerts our team, and you are only charged for successful responses. Errors are specific rather than generic: a nonexistent account is a 404 saying so by name, not a 503 dressed up as an outage; an unrecognised field is a 400 naming the field; a 502 or 503 means the source itself was unusable and is worth retrying after a wait.

Yes. All eight routes are MCP tools on the same FetchLayer MCP server, named instagram_user_profile, instagram_user_posts, instagram_user_reels, instagram_post_detail, instagram_post_comments, instagram_reels_feed, instagram_trending_searches and instagram_resolve_url. Tool calls use the same key and are billed exactly like REST.