I wanted to know which Instagram carousels get shared, not just liked, so I needed likes, comments and share counts for a couple of thousand posts. Here is how it went with HikerAPI.
What I did
Pulled top posts for about 30 hashtags (/v2/hashtag/medias/top, two pages each): roughly 1,800 posts, around 190 of them carousels (media_type 8).
Pulled recent posts per author (/gql/user/medias, 12 posts per request) to get each author’s normal like level. That lets me compare a post with its author’s baseline instead of looking at raw likes.
What worked
Carousel metadata is rich: likes, comments, media type, caption, cover and slide count in one response.
A reshare/repost counter is present for a part of the posts (about 200 of 2,100 in my sample). Share counts were the thing I actually wanted, and I did not find them in the other scraper I tried, which returned only likes and comments.
Responses are fast, the JSON is consistent, and per-request billing is easy to follow in the dashboard.
Things to know before you build on it
Plain Python urllib requests were rejected with 403 by Cloudflare until I added a User-Agent header.
Share and save counts are not returned for every post. Check coverage on your sample before relying on them.
The hashtag “top” endpoint returns mixed content, mostly recent clips. Carousels were only about 11% of the results, so you need several hashtags to collect enough of them.
The $2 trial credit lasted about 100 requests. A study of a few thousand requests needs a budget, and the minimum top-up is a bigger commitment than a small experiment needs.
Takeaway
For share-based analysis of carousels, HikerAPI is the tool where I found the data I was looking for. I would use it again for a larger run.
Top comments (0)