Compare

How GroupMixer compares

Most group tools do one thing well: a quick shuffle, one fair split, or a golf rotation. GroupMixer plans many rounds and keeps your rules at the same time.

ChatGPT, spreadsheets and solver libraries are not on the map: they can do almost anything, if you build and check it yourself.

Built from the public awesome-group-generators reviews and GroupMixer’s benchmark, both maintained by GroupMixer’s author. Corrections welcome.

Group tools placed by how many rules they respect and how well they plan several roundsRespects more of your rules →Plans more rounds →RandomizersClassroom toolsGolf & rotationschedulersGroupMixer
  • GroupMixer
  • Randomizers
  • Classroom tools
  • Golf & rotation schedulers

Each dot is one reviewed tool, placed by its rule and multi-round features. Hover or tap a dot for its name.

Rules respected and rounds planned, per tool
ToolKindRulesRounds
Picker Wheel Team PickerRandomizers1½ of 60 of 8
Wheel of NamesRandomizers0 of 60 of 8
Clever GroupsClassroom tools4 of 6½ of 8
Classroomscreen Group MakerClassroom tools2 of 60 of 8
grueprClassroom tools4 of 6½ of 8
Social Golfer OnlineGolf & rotation schedulers3 of 67½ of 8
TourneyCard Golf RandomizerGolf & rotation schedulers1 of 64½ of 8
Good-Enough GolfersGolf & rotation schedulers2½ of 63 of 8
GroupMixerGroupMixer6 of 68 of 8

What are you using now?

Pick one to see what it typically gets wrong, next to the GroupMixer result.

What are you using now?

Coming from a random picker

Random picker
Round 4 of 4AvBeClDeEmFiGiHuIn

11 repeat meetings

GroupMixer
Round 4 of 4AvBeClDeEmFiGiHuIn

0 repeats · 36 of 36 pairs met

  • Plans every round, not just one shuffle

    Set the rounds and GroupMixer plans them together, so people keep meeting someone new.

  • Respects the rules you care about

    Preferences and pinned people shape the result instead of being fixed by hand afterwards.

Feature by feature

A category gets a dot when at least one of its tools has the feature, which favours the other tools.

  • Yes
  • Partly
  • No
Multi-round rotations
GroupMixer8/8Randomizers0/8Classroom1/8Schedulers7½/8
Multiple roundsYesNoPartlyYes
Limit repeat meetingsYesNoNoYes
Maximise new contactsYesNoNoYes
Pair meeting targetsYesNoNoPartly
Who-met-whom matrixYesNoNoYes
Partial attendanceYesNoPartlyYes
Per-round rulesYesNoNoYes
Per-round group sizesYesNoNoYes
Constraints and balance
GroupMixer6/6Randomizers1½/6Classroom5/6Schedulers4½/6
Keep togetherYesPartlyYesPartly
Keep apartYesPartlyYesYes
Weighted preferencesYesNoYesPartly
Pin a person to a groupYesNoPartlyYes
Balance by attributeYesPartlyYesPartly
Minimum group sizesYesNoPartlyYes
Workflow and privacy
GroupMixer10/10Randomizers6½/10Classroom7/10Schedulers6½/10
Runs in the browserYesYesYesYes
No account neededYesYesPartlyYes
CSV importYesYesYesPartly
CSV exportYesYesYesYes
Zoom breakout CSVYesNoNoNo
API accessYesYesPartlyPartly
Result diagnosticsYesPartlyYesYes
Reproducible resultsYesNoNoPartly
Presentation modeYesYesYesNo
Edit groups after generatingYesNoYesYes
Where others are aheadWe list these on purpose.
GroupMixer0/3Randomizers3/3Classroom1½/3Schedulers1/3
Participants join via linkNoYesYesNo
Team leader assignmentNoYesPartlyYes
Spin or draw into groupsNoYesNoNo

Review data as of 1 August 2026 · dataset revision 552 · 503 tools reviewed in the public awesome-group-generators list.

Doing it yourself

ChatGPT, spreadsheets and solvers can express almost any rule. The difference is how much of the work is left to you.

GroupMixer

A dedicated group optimiser

2 steps for you

  1. Paste namesYou
  2. Pick rulesYou
  3. Solve and check every ruleGroupMixer
Checks every rule for you
YesBuilt-in evaluator checks every hard rule
Same input, same result
Yes
No account needed
Yes

Free for standard grouping · See pricingBenchmarks: 14 published cases

ChatGPT

Ask a chat assistant for groups

3 steps for you

  1. Write a promptYou
  2. Check every rule by handYou
  3. Re-prompt when one breaksYou
Checks every rule for you
NoYou verify the answer yourself
Same input, same result
Partly
No account needed
No

Paid Pro planBenchmarks: Not benchmarked

Excel

Build it in a spreadsheet

3 steps for you

  1. Build formulas or macrosYou
  2. Build your own checksYou
  3. Fix clashes by handYou
Checks every rule for you
PartlyOnly the checks you build
Same input, same result
Partly
No account needed
Partly

Microsoft 365 licenceBenchmarks: Not benchmarked

OR-Tools CP-SAT

Write your own solver model

3 steps for you

  1. Install the libraryYou
  2. Model every rule in codeYou
  3. Tune the solverYou
  4. SolveOR-Tools CP-SAT
Checks every rule for you
YesThe model enforces what you encodeRules left out of the model are not checked.
Same input, same result
YesDeterministic with a fixed seed and a single worker
No account needed
YesAn open-source library you install and run yourself.

Free, open sourceBenchmarks: Same 14 casesGroupMixer authored the published CP-SAT formulations; results are scored against the same scenario rules.

Hard cases, measured

The same scenarios, run in each tool and scored by one evaluator. The last case is one where OR-Tools does better.

32 people, 8 groups, 10 roundsPenalty: how often the schedule misses its goal of new contacts. Lower is better.

  • 0GroupMixerNobody shares a group twice
  • 330Good-Enough GolfersBest golf scheduler
  • 1,400OR-Tools CP-SATTuned solver model
Social Golfer 32×8×1032 people · 8 groups · 10 rounds
GroupMixer: 0 · perfect (best of 5 runs)GroupMixer0 · perfectGoogle OR-Tools CP-SAT: 1,400 (best of 5 runs)OR-Tools1,400Social Golfer Online: 385 (best of 5 runs)Social Golfer385TourneyCard Golf Randomizer: 440 (best of 5 runs)TourneyCard440Good-Enough Golfers: 330 (best of 5 runs)Good-Enough330
Social Golfer 32×8×10: best penalty, lower is better
ToolBest penalty
GroupMixer0 · perfect
Google OR-Tools CP-SAT1,400
Social Golfer Online385
TourneyCard Golf Randomizer440
Good-Enough Golfers330
Social Golfer 9–5–10 (45×9×10)45 people · 9 groups · 10 rounds
GroupMixer: 0 · perfect (best of 5 runs)GroupMixer0 · perfectGoogle OR-Tools CP-SAT: 2,935 (best of 5 runs)OR-Tools2,935Social Golfer Online: 781 (best of 5 runs)Social Golfer781TourneyCard Golf Randomizer: not supportedTourneyCardnot supportedGood-Enough Golfers: 627 (best of 5 runs)Good-Enough627
Social Golfer 9–5–10 (45×9×10): best penalty, lower is better
ToolBest penalty
GroupMixer0 · perfect
Google OR-Tools CP-SAT2,935
Social Golfer Online781
TourneyCard Golf Randomizernot supported
Good-Enough Golfers627
Six student workshop rotations120 people · 12 groups · 6 rounds
GroupMixer: 0 · perfect (best of 5 runs)GroupMixer0 · perfectGoogle OR-Tools CP-SAT: 576,576 (best of 5 runs)OR-Tools576,576Social Golfer Online: 95,095 (best of 5 runs)Social Golfer95,095TourneyCard Golf Randomizer: not supportedTourneyCardnot supportedGood-Enough Golfers: 103,103 (best of 5 runs)Good-Enough103,103
Six student workshop rotations: best penalty, lower is better
ToolBest penalty
GroupMixer0 · perfect
Google OR-Tools CP-SAT576,576
Social Golfer Online95,095
TourneyCard Golf Randomizernot supported
Good-Enough Golfers103,103
Social Golfer 196x14x14 (14-14-14)196 people · 14 groups · 14 rounds
GroupMixer: 750 (best of 5 runs)GroupMixer750Google OR-Tools CP-SAT: no valid resultOR-Toolsno valid resultSocial Golfer Online: not supported (≤120 people)Social Golfernot supported (≤120 people)TourneyCard Golf Randomizer: not supportedTourneyCardnot supportedGood-Enough Golfers: 4,412 (best of 5 runs)Good-Enough4,412
Social Golfer 196x14x14 (14-14-14): best penalty, lower is better
ToolBest penalty
GroupMixer750
Google OR-Tools CP-SATno valid result
Social Golfer Onlinenot supported (≤120 people)
TourneyCard Golf Randomizernot supported
Good-Enough Golfers4,412
Canonical station rotation 18×6×618 people · 6 groups · 6 rounds
GroupMixer: 30 (best of 5 runs)GroupMixer30Google OR-Tools CP-SAT: 0 · perfect (best of 5 runs)OR-Tools0 · perfectSocial Golfer Online: not supportedSocial Golfernot supportedTourneyCard Golf Randomizer: not supportedTourneyCardnot supportedGood-Enough Golfers: not supportedGood-Enoughnot supported
Canonical station rotation 18×6×6: best penalty, lower is better
ToolBest penalty
GroupMixer30
Google OR-Tools CP-SAT0 · perfect
Social Golfer Onlinenot supported
TourneyCard Golf Randomizernot supported
Good-Enough Golfersnot supported

Best penalty over repeated runs of each tool; lower is better, and 0 is a perfect schedule for that case’s objective. Penalties are comparable only within a case. Benchmark revision 18.

Every case vs OR-ToolsFull benchmark and methodology

Ratings and feature flags come from the awesome-group-generators methodology. Spotted something outdated or wrong? Request a correction.