Love Babbar's 450 DSA sheet, also called the DSA Cracker Sheet, is still a usable free problem list in 2026, but doing all of it is rarely the best use of your time. The copy we counted has 448 entries across 15 topics. About 400 of its links point to GeeksforGeeks rather than LeetCode, and around two dozen entries repeat a problem listed elsewhere in the sheet. It is worth it if you are preparing for campus placements or GeeksforGeeks-style coding rounds and have three months or more. For LeetCode-style interviews on a shorter timeline, the NeetCode 150 or Striver's sheets cover the core with less repetition. If you do use the 450, work a trimmed order of about 170 problems, laid out below.
Love Babbar shared the sheet as a spreadsheet in 2020. We counted the export bundled with the open-source 450dsa progress tracker on GitHub, which credits Love Babbar and links his original DSA Cracker Sheet. Copies drift by a few entries, which is why you will see 445, 448 or "450" quoted. Counts below are from October 14, 2026.
What is in the 450 DSA sheet
| Topic (sheet order) | Entries | Notes |
|---|---|---|
| Array | 36 | From reversing an array to the median of two sorted arrays |
| Matrix | 10 | Spiral traversal, search, rotation, maximum rectangle |
| String | 43 | Includes KMP, Rabin-Karp and Boyer-Moore pattern search |
| Searching & Sorting | 36 | Binary search on the answer (Aggressive Cows, Book Allocation) plus several SPOJ and HackerEarth problems |
| LinkedList | 36 | Singly, doubly and circular lists |
| Binary Trees | 35 | Traversals, views, construction, LCA |
| Binary Search Trees | 22 | Including merging two BSTs and the largest BST in a binary tree |
| Greedy | 35 | Classic scheduling problems plus six SPOJ problems |
| BackTracking | 19 | N-Queens, Sudoku, m-coloring, Knight's tour |
| Stacks & Queues | 38 | Many implement-it-yourself exercises |
| Heap | 18 | Heap implementation, merging K lists, running median |
| Graph | 44 | Core algorithms plus NP-hard problems such as the Travelling Salesman and Vertex Cover |
| Trie | 6 | Building a trie, shortest unique prefix, phone directory |
| Dynamic Programming | 60 | The largest section |
| Bit Manipulation | 10 | Set bits, power of two, power set |
| Total | 448 |
Three things stand out once you read the whole list rather than the topic headers:
- It is a GeeksforGeeks list. By our count, 400 links go to GeeksforGeeks articles or practice problems, 21 go to LeetCode, 11 to SPOJ and a handful to HackerEarth, HackerRank and other sites. In a spot check of 26 GeeksforGeeks and Techie Delight links in October 2026, every one still loaded, though most now redirect to newer addresses.
- It repeats itself. Kadane's algorithm appears three times (twice in Array, once in DP), and "rearrange characters so no two adjacent are the same" appears in String, Greedy and Heap. Edit distance, longest common subsequence, word break, merge K sorted lists and others appear twice. Strip the repeats and the three theory questions with no link, such as "Why strings are immutable in Java?", and you have roughly 420 distinct coding problems.
- Importance is marked by hand. Some entries carry notes like "[V.IMP]", and Kadane's algorithm gets "[V.V.V.V.V IMP]". There are no easy, medium or hard labels and no pattern groups.
Is the 450 DSA sheet worth it in 2026?
Where it still earns its place
- Breadth. Sixty DP problems, 44 graph problems and 43 string problems is more coverage than any 150-problem list.
- Placement-style rounds. If your placement tests or interviews use GeeksforGeeks-style problems and judges, the 450 maps to that style closely.
- Implementation practice. Building stacks, queues, heaps and graph algorithms from scratch is real practice for rounds that ask you to implement a data structure rather than call a library.
- It costs nothing. The sheet and the GeeksforGeeks practice problems it links to are free.
Where it falls short
- Size. At five new problems a day, 448 entries take about three months; at three a day, about five months. Most people never finish.
- Uneven difficulty. It mixes warm-ups like "find the maximum and minimum element" with problems almost no interviewer asks, such as the Chinese Postman problem, the Two Clique problem and counting all spanning trees.
- Not LeetCode-native. If your loops use LeetCode-style problems, you will spend time translating GeeksforGeeks statements and judges.
- Gaps in common patterns. There is no Longest Substring Without Repeating Characters, no Top K Frequent Elements, no Climbing Stairs or Decode Ways, and no Insert Interval. The sliding window family in particular is thin.
Our verdict: worth it as a source, not as a checklist. Use it to drill topics in depth, especially DP, graphs and linked lists, and skip the long tail.
How it compares with NeetCode 150 and Striver's sheets
| Love Babbar 450 | NeetCode 150 | Striver's SDE Pattern Sheet | Striver's A2Z DSA Sheet | |
|---|---|---|---|---|
| Problems | 448 entries, about 420 distinct | 150 | 179 | 495 |
| Structure | 15 topics, no difficulty labels | 18 topics in a fixed order; 28 easy, 101 medium, 21 hard | 17 modules split into pattern groups | 20 modules from beginner to advanced |
| Main judge | GeeksforGeeks (about 400 links) | LeetCode | LeetCode for 148 of 179 | LeetCode for 289 of 495 |
| Explanations | None built in; GeeksforGeeks articles and community solutions | A free video for each problem | Articles and videos for many problems | Articles and videos for many problems |
| Realistic time | 3–5 months in full | 6–10 weeks | 64.1 hours by takeuforward's estimate, before re-solves | 131.4 hours by takeuforward's estimate |
| Best for | Broad, placement-style practice with a long runway | LeetCode-style loops on a deadline | Algorithm depth once you know the basics | Learning DSA from zero |
The overlap is real but partial. By our manual read, roughly 70 of the NeetCode 150 appear in the 450 in the same or a near-identical form, spread across linked lists, trees, graphs, stacks and DP. Striver's A2Z sheet and his SDE Pattern Sheet share many of the 450's classics too, including job sequencing, minimum platforms, Aggressive Cows, Book Allocation, the celebrity problem and flattening a linked list, but link most of them to LeetCode or takeuforward's own articles. If you already finished one of those lists, the 450's added value is mainly its extra DP, string and implementation problems. Our comparison of Blind 75, NeetCode 150 and LeetCode 75 covers the LeetCode-native options in more detail.
A trimmed order: 172 problems in five phases
This is our own cut. It keeps each topic's interview staples, drops repeats and niche problems, and defers DP to the end so the earlier topics support it. At about 20 new problems a week, it takes nine to ten weeks.
| Phase | Topics | Keep | What to keep |
|---|---|---|---|
| 1 (weeks 1–2) | Array, Matrix, String, Searching & Sorting | 42 | Kadane's (once), minimize height difference, minimum jumps, duplicate in N+1, merge without extra space, merge intervals, next permutation, count inversions, stock buy and sell, longest consecutive subsequence, zero-sum subarray, trapping rain water, maximum product subarray, median of two sorted arrays; spiral, search, rotate, row-wise median, maximum rectangle; longest palindromic substring, KMP, Rabin-Karp, bracket reversals, valid IP addresses, both smallest-window problems, anagram groups, isomorphic strings, longest common prefix; first and last position, rotated search, square root, repeating and missing, majority element, pair with difference, four-sum, product array puzzle, k-th of two sorted arrays, Aggressive Cows, Book Allocation, Painter's Partition, minimum swaps to sort |
| 2 (weeks 3–4) | LinkedList, Stacks & Queues, Heap | 36 | Reverse, reverse in k-groups, detect, remove and find the start of a loop, add two numbers, intersection point, merge sort, middle, palindrome, flatten, clone with random pointer, n-th from end; balanced parentheses, O(1)-space min stack, next greater and next smaller, celebrity, postfix evaluation, largest histogram rectangle, longest valid substring, queue from stacks, stack from queues, LRU cache, petrol pump tour, rotten oranges, first negative in each window; heap implementation, k largest, window maximum, merge K arrays, merge K lists, smallest range, running median, connect ropes, reorganize string |
| 3 (weeks 5–6) | Binary Trees, Binary Search Trees | 28 | Level order, height, diameter, mirror, iterative inorder, preorder and postorder, left, right, top and bottom views, zigzag, balanced check, boundary traversal, build from inorder and preorder, LCA, distance between nodes, duplicate subtrees, k-sum paths; BST delete, successor and predecessor, validate, LCA, build from preorder, k-th largest and smallest, merge two BSTs, largest BST in a tree |
| 4 (weeks 6–8) | Greedy, BackTracking, Graph, Trie | 39 | Activity selection, job sequencing, fractional knapsack, minimum platforms, minimum coins, chocolate distribution, K negations; rat in a maze, N-Queens, Sudoku, m-coloring, palindromic partitions, combination sum, string permutations, K equal-sum subsets, remove invalid parentheses; BFS, DFS, cycle detection (directed and undirected), flood fill, islands, clone graph, word ladder, topological sort, finishing all tasks, alien dictionary, Dijkstra, Bellman-Ford, Floyd-Warshall, Prim, Kruskal, bridges, Kosaraju, bipartite check, cheapest flights within K stops; build a trie, shortest unique prefix, phone directory |
| 5 (weeks 8–10) | Dynamic Programming, Bit Manipulation | 27 | Coin change, 0/1 knapsack, subset sum, partition, LCS, LIS, edit distance, longest palindromic subsequence, word break, matrix chain multiplication, egg dropping, palindrome partitioning, unbounded knapsack, largest square of 1s, minimum cost path, maximum sum increasing subsequence, pair chain, interleaved strings, boolean parenthesization, weighted job scheduling, stock with k transactions, maximum sum rectangle; count set bits, two non-repeating elements, power of two, total set bits from 1 to n, power set |
What to skip or defer
- Repeats. Do Kadane's algorithm, edit distance, LCS, word break and merge K sorted lists once each, in the phase where they fit best.
- Theory prompts. The three unlinked questions are worth a minute of reading, not a problem slot.
- NP-hard and niche graph problems. Travelling Salesman, Vertex Cover, Chinese Postman, the Two Clique problem, counting spanning trees and counting triangles are interesting, but rarely asked as coding questions.
- Doubly and circular list variants. Rotating a doubly linked list in groups or splitting a circular list adds little once the singly linked versions are solid.
- Most SPOJ and HackerEarth entries. They are good contest practice but use input formats and constraints unlike interview problems. Keep Aggressive Cows, which is also on GeeksforGeeks.
- Warm-ups you already know. If reversing an array or finding the minimum and maximum takes you one minute, mark it done and move on.
How to work through it
- Pick one judge per problem. Solve on the GeeksforGeeks practice page the sheet links to, or on the LeetCode equivalent if your interviews will be LeetCode-style. Do not do both.
- Time-box. Give each problem 25–30 minutes before reading a solution, then re-solve it from a blank editor two days and a week later.
- Label the pattern. The sheet groups by data structure, not technique, so write the technique yourself: "binary search on the answer", "monotonic stack", "interval DP". Our 15 LeetCode patterns guide gives you a vocabulary, and the DP patterns guide is the companion for phase five.
- Explain out loud. Most interviewers ask you to talk through your approach and its complexity, not just submit code. CoPilot Interview's AI mock interview is free for 15 minutes a day: give it your résumé and a job description, weight the mix toward coding, and it reads each question aloud, scores your spoken answer out of 10 with what to fix and a model answer, and ends with a report card.
Practice explaining what the sheet taught you
Solving is half of a coding round; explaining is the other half. CoPilot Interview's free AI mock interview scores your spoken answers and shows a model answer, 15 minutes a day with no credit card.
Try a free AI mock interviewFAQ
How many questions are in Love Babbar's 450 DSA sheet?
The copy we counted in October 2026 has 448 entries across 15 topics, despite the name. About two dozen repeat a problem listed elsewhere and three are theory questions with no link, so the number of distinct coding problems is roughly 420.
Is the Love Babbar 450 sheet still worth it in 2026?
As a source, yes; as a checklist, usually not. Its DP, graph, string and implementation problems are still good practice, especially for placement-style rounds, but it is long, repeats itself and leans on GeeksforGeeks. A trimmed order of about 170 problems plus a few modern staples from the NeetCode 150 gets most of the value.
Love Babbar 450 vs NeetCode 150: which should I do?
NeetCode 150 if your interviews are LeetCode-style and you have about two months, since every problem is on LeetCode and has a free video. The 450 if you are preparing for GeeksforGeeks-style placement tests over three months or more and want broader coverage.
Love Babbar 450 vs Striver's sheets: which is better?
Striver's SDE Pattern Sheet is shorter (179 problems), grouped by pattern and mostly linked to LeetCode, and his A2Z sheet (495 problems) is built for learning from zero. The 450 has more raw breadth but no pattern grouping or difficulty labels. Many of the classics overlap.
How long does the 450 DSA sheet take?
All 448 entries take about three months at five new problems a day and about five months at three a day, before re-solves. The trimmed 172-problem order on this page takes nine to ten weeks at about 20 new problems a week.
Is the 450 DSA sheet free?
Yes. The sheet itself is free, and most of the problems it links to are free GeeksforGeeks practice pages or articles. A few links go to LeetCode, SPOJ, HackerEarth and HackerRank.