Loading game…

20×20 Nonogram Solver — Step-by-Step Solutions for Every Configuration

The 20×20 Nonogram Solver processes any valid 20×20 Japanese crossword, Griddler, or Picross clue configuration and returns the complete solution — or, for puzzles requiring hypothesis testing, the step-by-step logical path through which the solution is reached. The solver applies the same constraint-propagation and hypothesis-selection algorithms used by advanced human solvers, making it a precise and transparent tool for both unblocking specific puzzles and understanding the logical structure of 20x20 configurations.

How to Use the Solver

Step 1 — Enter your clues: Input the clue sequences for all 40 lines of your 20×20 puzzle — the row clues in order from top to bottom, and the column clues from left to right. Each clue is entered as a space-separated sequence of numbers (e.g., "3 2 4" for a three-block clue). The solver validates each entry against the line length to catch input errors before processing begins.

Step 2 — Run the solver: Click Solve. The solver processes your 400-cell configuration through its constraint-propagation and hypothesis-selection algorithm and returns the complete solution grid, with each cell clearly marked as filled or empty.

Step 3 — Review the solution path: The solver displays not just the final solution but the step-by-step logical path used to reach it — which lines were resolved in which order, which constraints forced each cell confirmation, and (for Expert through Evil configurations) which hypothesis cells were selected and what cascade they produced. This step-by-step output is the primary learning tool the solver provides.

Step 4 — Return to playing: Use the solution path to identify where your own solving approach diverged from the optimal path, then return to the puzzle and continue from your current position — or start a fresh puzzle with improved technique. The solver is designed to accelerate your development as a 20x20 solver, not to replace the solving experience.

How the 20×20 Solver Algorithm Works

Initialisation phase: For a 20×20 grid, the solver initialises 40 line constraint sets across 400 cells using a parallelised initialisation architecture. Lines are grouped into four initialisation batches; each batch is fully initialised (enumeration, overlap, segment analysis) before the confirmed cells from that batch are propagated to all intersecting lines in the next batch. This batched approach reduces total initialisation time by exploiting the information flow between batch-adjacent lines.

Constraint propagation phase: Propagation across the 40-line network uses a quadrant-aware cascade architecture. The 20×20 grid is divided into four quadrants; cascade chains are tracked by quadrant origin and destination, and lines in cascade-receiving quadrants are prioritised in the propagation queue. This quadrant awareness prevents high-priority cascade information from being processed after lower-priority lines in non-cascade quadrants, reducing total propagation rounds by 20 to 35 percent compared to non-quadrant-aware approaches.

Hypothesis resolution phase: For 20×20 hypothesis configurations, the solver applies a two-phase hypothesis selection process. Phase 1 identifies all cells in lines with two or fewer remaining arrangements — these are the highest-leverage candidates. Phase 2 simulates a three-step propagation from each Phase 1 candidate under both assumptions and ranks candidates by three-step cascade yield. The highest-ranked candidate is selected as the hypothesis target, consistently producing cascades that resolve 40 to 60 percent of remaining ambiguities per cycle.

Accuracy and Reliability

The 20×20 solver is guaranteed to find the unique solution to any well-formed 20×20 nonogram — a puzzle constructed so that exactly one cell configuration satisfies all clue constraints simultaneously. For puzzles with ambiguous clue sets (where multiple valid solutions exist), the solver identifies the ambiguity and reports which cells have multiple valid states rather than selecting arbitrarily among valid solutions.

All solutions returned by the solver are verified against the complete clue set before display — ensuring that the solution reported is always valid, never partial, and never the result of an incorrect hypothesis branch that was not properly resolved.

When to Use the Solver

The solver is most valuable in four specific situations:

Blocked at a specific point: You've applied every technique you know to every 20x20 line and cannot identify the next confirmed cell. The solver identifies the exact next deduction — whether a standard elimination or a hypothesis step — and explains why it follows from the current constraint state.

Learning hypothesis technique: You're developing hypothesis-and-verify skills and want to compare your hypothesis selection to the solver's. The solver's hypothesis target, assumption direction, and cascade sequence provide a concrete benchmark for evaluating your own selection strategy.

Verifying a partial solve: You want to confirm that your current grid state — with some cells already confirmed — is consistent with the unique solution before investing further time in the puzzle.

Post-solve analysis: You've completed the puzzle independently and want to understand whether the path you took was optimal — or whether there was a shorter sequence of deductions that would have reached the same solution in fewer steps.

Play 20×20 Nonograms

Ready to put the solver's insights into practice? The 20×20 puzzles are available across all six difficulty levels:

20×20 Easy20×20 Medium20×20 Hard

20×20 Expert20×20 Extreme20×20 Evil

FAQ

Yes — the solver processes any valid 20×20 clue set where the sum of each row's clue values (plus minimum gaps) does not exceed 20 cells, and similarly for columns. Clue sets that violate these constraints are flagged as invalid before processing begins.

No — the solver operates independently of your puzzle session. Entering your clues into the solver and reviewing the solution does not modify your in-progress puzzle. You can return to your puzzle at any stage and continue from exactly where you left off.

Yes — the solver handles all difficulty tiers including Evil, which requires nested hypothesis trees. The solver's hypothesis-selection and cascade-propagation algorithms are designed specifically to handle the deep conditional reasoning that Evil 20x20 configurations require, and the 20×20 solver resolves hard configurations in under thirty seconds, expert in under ninety seconds, extreme in under four minutes, and evil configurations in under ten minutes.

If the clue configuration you enter has no valid solution — either because of an input error or a genuinely infeasible clue set — the solver reports this explicitly rather than returning an incorrect partial result. Check your clue entries against the original puzzle to identify any transcription errors.