Three steps, no account
Name the giveaway, paste the comments, press draw. The name is there because most people running these do it every few weeks β it lands on the result and in the announcement text you copy out, so a month later you still know which draw was which.
Nothing is uploaded. The parsing and the draw both happen inside your browser, which is why it copes with tens of thousands of comments without a loading spinner, and why closing the tab wipes everything.
Wheel or reel
The name reel flickers through handles and stops on one. It is quick, and it is what you want when you are drawing five winners and do not need a show.
The wheel is the one worth recording. It spins for about five seconds and slows onto the winner, which fits neatly into a Story or a Reel. One honest limitation: a wheel with two thousand names on it is unreadable, so the wheel displays twelve. The winner is picked from the complete pool first, using every single entry, and then shown on the wheel β the wheel is the presentation, not the lottery. The note under it says so on the page too, because a giveaway tool that quietly fudges this is worse than no tool.
What it checks for you
- Duplicate handles. Somebody who commented forty times gets one entry, unless you deliberately switch to counting every comment.
- Tag requirements. If the rule was "tag two friends", comments with fewer mentions drop out. Tagging the same friend twice counts once.
- Keywords. Require "done" or "ikutan" and the drive-by "nice" comments stop competing with real entrants.
- Your own accounts. Add the brand account and your staff to the never-draw list once and forget about it.
- Backups. Drawn in the same round as the winners, because a good share of winners never reply.
Everything filtered out is listed with the reason it was skipped, so you can tell an unusual comment format apart from an actual rule breach before you draw.
Why the seed matters
Any tool can produce a random name. The problem is the reply under your announcement saying you picked your friend. A screen recording only proves that something happened on your screen.
This draw comes from a seed that gets printed on the stub along with a fingerprint of the entry list. Post those two values with the result and anyone can paste the same comments, type the same seed, and land on the same winner. The doubt has somewhere to go other than your DMs.
Questions
Does it work with Threads?
Yes. Threads replies copy out in much the same shape as Instagram comments, and the parser treats them identically.
How do I get thousands of comments out of a post?
For your own post, the Instagram Graph API returns them as JSON, which you can paste straight in. Otherwise, expand every comment on the web version and copy the block. The full guide walks through both.
Is it really random?
The seed comes from your browser's cryptographic random number generator, and every entry in the pool has an equal chance. Being reproducible afterwards is not the same as being predictable beforehand β you cannot know the seed until it exists.
Can a company use this for a client's campaign?
Yes, and the giveaway name plus the copied announcement is meant for exactly that: something you can paste into a report showing which draw was run, when, over how many entries, and how the client can verify it.
Does it cost anything?
No. There is no account, no limit on comments, and no watermark on the result.