Drag and Drop
The Drag and Drop content type generates an activity where students drag elements to their correct place, in one of two modes: labeling an image or categorizing elements into groups. It is one of the eight H5P types that respect the Generate images option of the AI modal — and in labeling mode the image is mandatory.
AI-generated configuration
| Setting | Description | Possible values | Default |
|---|---|---|---|
| Mode | How the activity is set up. | Labeling (drag labels onto a generated image) / Categorization (drag elements into groups) | Inferred from the prompt |
| Number of labels (labeling mode) | How many draggable labels the activity contains. | Any number — no cap; an exact requested count (e.g. 6 or 15) is generated exactly, without trimming | 3 to 8 labels when no count is requested |
| Number of groups (categorization mode) | How many categories the elements are classified into. | Any number — no cap; an exact requested count is respected | 2 to 4 groups when no count is requested |
| Elements per group (categorization mode) | How many draggable elements each category holds. | At least 2 elements per group, always | At least 2 per group |
| Error penalty | Whether wrong drops subtract points. | No penalty — the student keeps the partial score | No penalty |
Generated content
- Labeling mode: an AI-generated image of the subject, plus draggable labels with their drop zones placed coherently with what is visible in the image; each zone is verified against the generated image.
- Categorization mode: the requested groups, each with at least two elements, every element assignable to its correct category, over a background image or a neutral panel.
Rules and behavior
- No error penalty: the student keeps the partial score for correct drops. Only the variant embedded in a Question Set penalizes errors.
- Labeling mode requires the image: run it with Generate images on — without the image this mode cannot be produced.
- Labels that cannot be placed with confidence are discarded rather than left on a wrong zone: the activity may ship with fewer labels than requested before shipping misplaced markers.
- Neutral-panel fallback (categorization mode): if the background image fails, the activity is generated all the same over a neutral panel.
- Neither mode has a cap: exact requested counts of labels (e.g. fifteen) or groups (e.g. three, with three elements each) are generated exactly, without trimming; each group always keeps at least two elements.
- With a syllabus or source document uploaded, the labels, elements, and categories are terms taken strictly from the document, with no invented components.
Example prompts
Without syllabus
Produces a labeling-mode activity with five draggable labels over the generated image, drop zones coherent with the image, and no error penalty:
I need an H5P drag and drop activity for the industrial safety course. Show a worker with their personal protective equipment and have the student drag the labels helmet, gloves, goggles, boots, and vest to the matching body part.
Produces a categorization-mode activity with three groups, each with at least two elements and partial credit with no penalty:
Prepare an H5P drag and drop for the logistics course where the student classifies goods by their storage type: refrigerated, dry, and hazardous. Put three or four example products in each category.
Produces a labeling-mode activity with exactly fifteen labels, since labeling has no cap:
I need an H5P drag and drop to label fifteen bones of the human body on a skeleton: skull, clavicle, sternum, humerus, radius, ulna, pelvis, femur, patella, tibia, fibula, scapula, ribs, spine, and jaw.
With syllabus
Produces a categorization-mode activity with exactly three groups and three elements per group, all taken strictly from topic 4 of the uploaded document:
Build an H5P drag and drop about topic 4 of the document I uploaded with exactly three categories taken from that topic. In each category, three elements the student must drag to their group.