It was the last working day of the month, the day I send my clients their Instagram reports. I had a coffee, a half-finished report and a script that had worked fine the previous month. Then it died on a login challenge, and I spent the next hour re-authenticating an account instead of sending anything. I'm a freelance social media manager who writes just enough Python to get into trouble, and that hour was when I started asking whether I should be maintaining this at all.
The job
Once a month I pull stats for each client's account (followers, recent posts, how the posts did) and turn them into a report. The volume is small, but the deadline is real and the numbers have to be right, because clients read them.
instagrapi: what it's good at
I used instagrapi first, and I'd still recommend it in the right situation. It's free, it's well maintained, and it gives you a lot of control. For a hobby project or a one-off script it's hard to beat.
My problems were all around the login:
- It works through a logged-in account, so I had to manage sessions.
- When Instagram flagged the session, I got challenges and, once, a locked account.
- I never felt safe using an account that mattered for client work.
None of that is the library's fault. It's how that approach works.
The hosted API route
I moved the collection step to HikerAPI, a hosted REST API. You send your key in an x-access-key header and call an endpoint. It's pay-per-request, starting at about $0.001 per request, and there are 100 free requests to try it. I used those to compare the output with what I see in the app before changing anything.
Here is the call for one profile:
import requests headers = {"x-access-key": "YOUR_KEY"} r = requests.get("https://api.hikerapi.com/v2/user/by/username", params={"username": "nike"}, headers=headers) print(r.json()) That's the whole login flow: there isn't one. My report script went from managing sessions and retries to looping over a list of usernames and saving each response to a file, so a failed run only re-fetches what's missing.
Side by side
| instagrapi | Hosted API | |
|---|---|---|
| Upfront cost | Free | Free tier to test |
| Ongoing cost | Accounts, maybe proxies, your time | Per request |
| Login/session handling | You | Provider |
| Ban or lock risk | On your account | Not on your accounts |
| Control over how data is fetched | High | Low |
| Setup effort | Medium | Low |
| Dependency | The library and Instagram | The provider's uptime |
When I'd pick each
Use instagrapi if:
- you want full control and are happy to maintain sessions,
- you're doing a hobby project or a one-off,
- you have throwaway accounts and proxies already.
Use a hosted API if:
- you have a deadline (mine is monthly) and can't afford a broken run,
- you don't want a real account tied to the job,
- your volume is low enough that per-request pricing beats paying for idle infrastructure.
What to watch for
- Cost at scale. Per-request pricing is cheap for a few clients. If you pull a very large number of records, do the math first.
- Correctness. Spot-check results against the app before sending a report to a client. That applies to any source, including your own scraper.
- Dependency. If the provider goes down, so does your pipeline. I keep the saved responses from each run as a fallback.
Where I ended up
For my use, small and deadline-driven, the hosted API removed the part that kept breaking. If you're a developer who enjoys controlling every layer, or you run a lot of accounts yourself, instagrapi may still be the better fit.
Disclosure: the provider offers free credits for write-ups like this one, so weigh my opinion accordingly. The comparison above comes from what I actually ran into.
What's your setup for recurring Instagram reports? Still running your own, or have you handed it off?
