What is a random group maker?
A random group maker splits a class list into teams so you do not have to count off, draw names from a hat, or let students pick — which always leaves somebody last. You paste the names, say how big the groups should be or how many you want, and it deals them out fairly.
This one runs in your browser with no account and no student logins. It also handles the things that make a grouping actually usable in a classroom: keeping two students apart, remembering who has already worked with whom so partners change over a term, rotating the roles inside each group, and leaving absent students out for the day without losing them from the class list.
Even groups, and why that is harder than it sounds
The obvious way to split a list is to shuffle it and cut it into chunks of the size you asked for. It is one line of code and it is wrong, because everything left over lands in the final chunk. Twenty-five students in groups of four becomes six groups of four and one group of one — a child sitting on their own while everyone else works in a four.
This tool used to do exactly that, and the page used to describe it as "the last group simply gets the remaining students — no one is left out", which was not true. Students are now dealt round-robin, one at a time around the groups, so the sizes come out 4,4,4,4,3,3,3 and never differ by more than one. The same applies when you ask for a number of groups instead of a size: seven groups from thirty students gives five fours and two fives, not six fours and a six.
Both ways of splitting are here because teachers need both. "Groups of three" is what you want when the activity needs three roles; "six groups" is what you want when there are six tables. With only one of them, you end up doing the division in your head and hoping the remainder falls somewhere sensible.
The rules are the point
Every teacher has two or three students who cannot be in the same group. That is the actual reason grouping gets done deliberately rather than by counting off, and it is the thing a random generator is most likely to get wrong — if it seats them together, the class assumes you did it on purpose.
Rules here are plain text, one per line. Commas keep people apart, a plus sign keeps them together: "Jake, Marcus" and "Priya + Sam". There is no roster to build first and no menu to work through, which matters because the whole appeal of this kind of tool is that it takes twenty seconds. Names that are not in the class list are reported rather than silently ignored.
When the rules cannot all hold at once — three students who must all be apart, but only two groups — the arrangement says which rule it broke, by name. That is the difference between a constraint feature you can rely on and one you have to check by eye every time.
Who has already worked with whom
Random grouping every week does not actually mix a class. With twenty-five students in fours there are three hundred possible pairs and only thirty-three of them happen each week, so by chance alone the same people keep landing together while others never meet. Teachers notice this and start intervening by hand, which is how you end up back at fixed groups.
So this one remembers. Press keep after you read the groups out and the pairings are recorded for that class; the next grouping then works to avoid partners who have already worked together, weighting repeat offenders more heavily than first-time pairs. Over eight simulated weeks with a class of twenty-five, that took repeated pairings from seventy-six down to eighteen. It compares against every grouping you have kept, not just the previous shuffle.
You can also see it. The grid shows every student against every other one: an empty square is a pair who have never worked together, and the darker the square the more often they have. It is the quickest way to find the two students who somehow keep ending up on the same table — and to prove to yourself that the mixing is actually happening.
Roles, absences, and getting it onto the desks
Roles inside a group — Leader, Scribe, Reporter, Timekeeper — stop the first thirty seconds being a negotiation. Handing them out once is easy; rotating them is the useful part. Here the tool counts who has held what, so the confident child is not Leader every week and the quiet one gets a turn without you having to engineer it. Marking someone absent leaves them out of today's groups while keeping them in the class and its history, which is different from deleting the name and losing everything the tool knows about them.
When the groups are set, the list prints for your desk and the table tents print one page per group — fold them, stand them on the tables, and the names and roles are visible from both sides. From there the seating chart generator arranges the room those groups sit in, and the classroom timer runs the activity itself.