---

name: django-pagination-performance description: Improve Django and Django REST Framework pagination for large querysets using bounded page sizes, stable ordering, indexed ordering, offset limits, keyset pagination, DRF CursorPagination, and count-query tradeoffs. Use when list pages, admin views, APIs, exports, or infinite scroll become slow, inconsistent, or memory-heavy as result sets grow. license: MIT compatibility: Codex, Claude Code, and other Agent Skills-compatible clients. metadata: version: "0.1.0" displayName: Django Pagination Performance category: Django tags: django,pagination,performance,querysets,drf

Django Pagination Performance

Use this skill when a Django list view or API slows down as pages get deeper or result sets grow. Pagination is a query-design problem as much as a response-shaping problem.

Workflow

1. Confirm the list contract.

   • Is arbitrary page access required, or only next/previous?
   • Does the UI need a total count?
   • Can ordering be fixed and stable?
   • What page size limit is acceptable?

2. Measure the current query.

   • Capture SQL for the page query and count query.
   • Check ordering, indexes, and high page numbers.
   • Use QuerySet.explain() for deep pages.

3. Choose the pagination style.

   • Use Django Paginator for moderate result sets and arbitrary page access.
   • Use capped offset pagination when page numbers are useful but deep pages should be limited.
   • Use keyset/cursor pagination for large feeds, timelines, logs, and infinite scroll.
   • Use DRF CursorPagination for API next/previous navigation with stable ordering.

4. Make ordering deterministic.

   • Use a unique or nearly unique immutable ordering field.
   • Add a primary-key tie breaker when needed.
   • Ensure the index matches filters plus ordering.

See pagination-patterns.md for Django and DRF examples.

Safety Notes

• Deep LIMIT/OFFSET pages can be slow because the database still walks skipped rows.
• Unordered querysets produce inconsistent pages.
• Cursor pagination restricts arbitrary page jumps and user-controlled ordering.
• Large exact counts can dominate list latency.

Verification

Measure first page, representative deep page, count query, and insertion/deletion consistency for the chosen pagination contract.