Bingo cards are simple grid sheets that let players mark matching calls until a row, column, diagonal, or another agreed pattern is complete. Outside regulated number bingo, the same format works well for classroom vocabulary drills, workshop observation games, team retrospectives, and party prompts.
This generator turns your own prompt list into a randomized square board and a printable card. You can size the grid from 3 x 3 to 10 x 10, reserve the middle square on odd grids, add column headers, change the theme and layout preset, and place a title, subtitle, or footer note directly on the card.
It is built for batches as well as one-off sheets. A single run can create up to 100 cards, each tied to the current seed and settings, so the same setup can be reproduced later when you need matching materials for a class, a workshop series, or a second event host.
A common use case is a standup or retro session with 24 unique prompts on a familiar 5 x 5 board that keeps the middle square reserved, adds bold headers, and prints 12 variations for the room. The card preview shows both the current card number and the active seed, which makes reruns practical instead of guesswork.
The main boundary is that this tool is a layout generator, not a regulated gaming system. It removes exact duplicate prompt lines after trimming, but if the unique pool is smaller than the number of cells required for the chosen layout, it deliberately repeats entries to fill the board. It also does not enforce pack-wide uniqueness, serial tracking, caller logic, or prize-game compliance.
Start with the prompt list, because that decides whether the board will feel varied or repetitive. For a familiar classroom or party layout, 5 x 5 with the reserved middle square works well because it needs 24 unique prompts and keeps the card easy to scan. Smaller boards are better when the group will move quickly or you only have a short list of ideas.
If your source list is rough, generate once before exporting anything and read the Summary tab. Duplicates trimmed tells you repeated source lines were collapsed before shuffling, and Items short tells you whether the generator had to recycle prompts to fill the board. When Items short is above zero, repeated entries are expected rather than a bug.
Use a seed when you need the same pack again or when two people need to print matching sets. Leave the seed blank when you want a fresh shuffle. If you like the current result, the Use current button copies the active seed back into the input so you can preserve the exact setup.
This tool is a good fit for custom text bingo, icebreakers, workshop checklists, and observation games where the organizer controls the prompts and the calling rules. It is a poor fit for licensed or audited bingo operations that require fixed card standards, formal verification, or stricter control over duplicate cards across a pack. Before you export, make sure Items short is zero if you want to avoid repeated prompts caused by a short source list.
The generator begins by normalizing the prompt source. Line endings are unified, tabs and commas are treated as separators, blank lines are ignored, and each remaining line is trimmed. Exact duplicates are removed with no case folding, so Bug fix and bug fix are treated as different prompts while repeated copies of the same trimmed line collapse into one entry.
Board size is clamped to 3 through 10 on each side. The number of prompts actually needed depends on whether the reserved middle square is active. On odd-sized grids, reserving the middle square subtracts one prompt slot; on even-sized grids, that option has no effect because there is no single middle cell.
Randomization is reproducible rather than opaque. An explicit seed is interpreted either as hex or as text hashed to a 32-bit value. If you leave it blank, the tool generates a random 32-bit base seed. Each card then derives its own pseudo-random stream from that base seed, the card index, and the shuffle count, so the same seed and settings reproduce the same pack while a new shuffle changes it.
The tool does not stop when the prompt pool is too small. Instead, it shuffles the unique list, cycles through it until every required cell is filled, and then builds the grid. That is why Items short matters so much in the Summary tab. It predicts repeated prompts before you print. The package has no tool-specific backend or Lambda path, but current state is URL-backed, and document export loads its helper script the first time you request a DOCX file.
The required prompt count is determined by grid size and the middle-square rule:
| Symbol or field | Meaning | Range or type | Where it appears |
|---|---|---|---|
n |
Grid side length | integer 3-10 | Grid size, Summary |
m |
Middle-square offset | 0 or 1 | Derived from Centre tile |
C |
Prompts required to fill one board | integer | Cells needed |
Unique items |
Prompt count after trimming and exact deduplication | integer | Summary |
Duplicates trimmed |
Repeated source lines removed before shuffle | integer | Summary |
Items short |
How many prompt slots exceed the unique pool | integer | Summary |
| Stage | Package behavior | Visible consequence |
|---|---|---|
| Input cleanup | Split on line breaks, tabs, and commas; trim blanks; remove exact duplicates | Unique items can be lower than raw source lines |
| Header handling | Use custom labels when present; otherwise default to B I N G O for 5 x 5 and sequential letters for other sizes |
Short header lists are padded rather than rejected |
| Short prompt pools | Cycle through the shuffled unique list until every required cell is filled | Repeated prompts appear on the board when Items short is above zero |
| Batch generation | Derive a separate pseudo-random stream for each card from the base seed and card index | The same seed reproduces the same set when settings stay fixed |
| Exports | Capture the rendered card view for PNG, or capture every generated card for PDF and DOCX page composition | PNG saves the current card, while PDF and DOCX cover the full set |
Item Coverage is also narrower than it may look at first glance. The notes are calculated from the currently viewed card only, not across the entire batch. That makes the tab useful for checking repeats on one board, but not for proving pack-wide uniqueness.
Follow this sequence when you want a predictable card pack instead of a quick experiment.
Title, Subtitle, and Footer note if the cards need event labels or instructions. These appear directly on the printable sheet.Prompts, one entry per line. Watch the count below the textarea so you can see both the unique total and any exact duplicates removed from the source.Design and set Grid size, Centre tile, Column headers, Colour theme, and Layout preset. If header labels look odd later, come back here and match the number of header labels to the grid size.Advanced and Export next. Set the Seed if you need repeatability, then adjust Cell size, Font scale, Cards to build, Cards per page, Page size, Orientation, and Margins.Generate cards. Use Viewing card, Prev, Next, or Shuffle to inspect the set. If the Summary tab shows Items short above zero, stop and add more unique prompts before printing if you want to avoid repeated entries caused by a short pool.Item Coverage, then choose the export that matches your goal. Export PNG saves the current card only, while Export PDF and Export DOCX package the whole generated set according to the current page layout options.The fastest truth check is the trio of Unique items, Cells needed, and Items short. If Items short is zero, the card can be filled without shortage-driven repeats. If it is above zero, repeated prompts are expected because the board needs more cells than the unique source pool can supply.
Duplicates trimmed means exact repeated source lines were collapsed before shuffling. If you expected repeated lines to act like weighting, this tool will not preserve that behavior.Seed identifies a reproducible setup, but random-looking cards do not prove pack-wide uniqueness or regulated fairness. Verify the printed pack and your calling rules if the event is more formal than a classroom or social session.Item Coverage describes only the card currently on screen. Use it to inspect one board, not to certify the whole batch.Before exporting, the practical next check is simple: make sure the current card looks right, Items short matches your tolerance for repeats, and the displayed seed is the one you want to keep for future reruns.
A facilitator pastes 24 unique retrospective prompts, keeps a 5 x 5 grid with the reserved middle square, sets Cards to build to 12, and enters a memorable seed. The Summary tab shows Unique items 24, Cells needed 24, and Items short 0, which means the board does not need shortage-driven repeats. Reusing the same seed and settings later recreates the same 12-card pack.
A teacher enters 10 unique vocabulary prompts for a 4 x 4 board with no reserved middle square. The generator reports Cells needed 16 and Items short 6. When repeated words appear on the card, that is expected behavior rather than a failed shuffle, because the tool cycles through the shuffled prompt pool until all 16 cells are filled.
An event host builds 20 cards, checks Viewing card to inspect a few layouts, and then clicks Export PNG. The saved file contains only the currently displayed card, because PNG captures the visible board. The corrective path is to use Export PDF or Export DOCX, which capture every generated card and arrange them according to Cards per page, Page size, and Orientation.
Yes. Keep the same Seed, prompt list, grid size, centre-tile setting, headers, style options, and card count. Change any of those and the generated pack will diverge even if the seed text stays the same.
Because the board can need more prompt slots than the unique pool provides. Check Items short in the Summary tab. When that number is above zero, the generator reuses prompts to fill the missing cells.
No. Exact trimmed duplicates are removed before shuffling, which is why Duplicates trimmed can be greater than zero. If you need weighting, you would have to vary the prompt text itself, because repeated identical lines collapse into one entry.
PNG saves the current visible card. PDF and DOCX capture every generated card and place them into pages using the current Cards per page, Page size, Orientation, and Margins settings.
The package has no tool-specific backend or Lambda handler for prompt processing. Text dropped into the page is read in the browser, and exports are rendered locally. The privacy caveat is that current state is URL-backed, so the active list and settings may be recreated from the page link, and the DOCX helper script is fetched the first time you request that export.