Custom Puzzle Data

You define your game mechanics. We generate data that works for those mechanics.

Anagram Puzzles

For letter-tile games, anagram challenges, and Wordle-style spelling games. We provide a set of letter tiles, target words with optional semantic clues, and a complete list of valid anagrams and other letter combinations—everything a player might legitimately form from the available letters.

Puzzle Generation

Clue Generation

Pathfinding Puzzles

For semantic navigation games where players traverse meaning-space—connecting concepts through chains of associations. We provide validated origin-target pairs solvable in exactly 3 moves (4 hops through meaning-space), with multiple solution paths, precalculated hints, and backward “convergence” clues that tell players when they’re getting close.

Puzzle Generation

Hint Generation

Why hints must be precalculated: Finding valid hints requires enumerating all paths—not just one—to guarantee every hint actually reaches the target. With 15 words visible at each hop, that’s 50,000+ candidate paths per puzzle. Ranking hints by how many paths use them (more paths = more downstream options) requires global knowledge. And compressing hint data for mobile delivery (74% size reduction via ID encoding) requires knowing all hints upfront. Runtime computation would add seconds of latency; precalculation makes hints instant and guaranteed correct.

Themed Categories

For NYT Connections-style games with intentional decoys—words that tempt players into wrong groupings, where ambiguity is the challenge. Or for other category games that need mutually exclusive pools where no word could plausibly belong to two groups.

Clue-Based Puzzles

For crossword-style games and semantic guessing games—short clues designed for gameplay, plus related words that don’t give away the answer morphologically. Players identify a target word from semantic clues, but there might be decoys that match some (not all!) of the clues.

Clue Selection

Decoy Selection

Why decoys work: Each decoy must match 2–3 of the 4 clues—enough to seem plausible—but have at least one “zero” (a clue it cannot possibly connect to). Players eliminate decoys by finding the impossible connection. The answer wins by having no zeros: it connects to all 4 clues through different senses of the word.

What You Get

We deliver production-ready puzzle data for all your levels—hundreds or thousands, depending on your game’s needs. The data is generated exclusively for your game, unique to your studio. Every puzzle draws from Linguabase’s complete vocabulary stack: vocabulary rankings, semantic associations, sense clouds, and more.

• Compact data payload—minified JSON optimized for your game client, modular so you can update or swap level packs without redeploying
• Human-readable version—uncompressed, commented format you can inspect to understand the data structure and verify puzzle quality
• Clear documentation—schema reference, field explanations, integration examples
• Embedded difficulty scores—each puzzle tagged with difficulty metrics so your game can dynamically adjust challenge to individual players
• Content filtering—pre-vetted for your audience (family-friendly, teen, adult)

Already generating your own puzzles? License the underlying data → to improve what your content team produces.

The four formats above are examples. Your game has its own mechanics—we’ll design a data structure that fits, exclusive to you.

Get in Touch

We can generate test datasets for creative development and playtesting—small batches to validate that your mechanics work. When you’re ready, we produce data for thousands of production levels.

We’ve been generating word puzzles about meaning for over a decade. The edge cases are where all the work is.

Describe your game mechanics—how players interact, what constitutes a “level,” what data you need at runtime—and we’ll design a puzzle data format that plugs directly into your game.