Three-Tier Correction Engine
Every transcription passes through three matching tiers. Only word-boundary matches are applied — partial-word replacements are prevented.
Exact & Phrase Lookup
O(1) HashMap lookup for single-word corrections and multi-word phrase matching. Fastest tier — handles known misspellings and abbreviation expansions instantly.
k8s→KubernetesSymSpell Fuzzy Matching
Edit distance-based matching (configurable 0-3 distance, default 2) using prefix-based indexing. Catches typos and mishearings while a spell validator gate prevents correcting valid English words.
tenserflow→TensorFlowPhonetic Matching
Catches homophones and similar-sounding words that edit distance alone would miss. Essential for voice dictation where words are heard, not typed.
their/there/they're→Correct form based on contextPrompt Words — Vocabulary Boosting
Prompt words go deeper than post-processing — they bias the transcription engine itself toward specific vocabulary before any text is produced.
For the Whisper backend, prompt words are passed as "previous context" so the model expects these terms. For Parakeet, they're compiled into CTC vocabulary models that boost probability at the acoustic decoder level.
Use Cases
- Brand names and product names (e.g., "Kubernetes", "PostgreSQL", "Next.js")
- Personal names and team member names
- Medical or legal terminology
- Company-specific acronyms and jargon
- Foreign words used frequently in your workflow
Dictionary Management
Categories
Organize corrections by category (medical, legal, programming, names) for easy management.
Usage Tracking
See how often each correction is applied. Identify which entries are most valuable.
Import & Export
Import and export your dictionary as JSON. Share word lists across devices or teams.
Dictionary Packs
Premium bundled correction databases with per-pack enable/disable and automatic version updates.
Frequently Asked Questions
What are prompt words?
Prompt words bias transcription toward specific vocabulary — proper nouns, technical terms, brand names. For Whisper, they're passed as 'previous context'. For Parakeet, they're compiled into CTC vocabulary models that boost probability at the acoustic decoder level.
Will the dictionary change correctly spelled words?
No. The spell validator gate ensures that SymSpell fuzzy matches are validated against a spell checker — valid English words are not 'corrected' by fuzzy matching. Only genuinely misspelled or misheard words are fixed.
Can I see what was corrected?
Yes. In the live preview, corrected words are shown with gradient color highlighting. Click any highlighted word to see the original transcription before correction.
How do prompt words differ from dictionary entries?
Dictionary entries fix text after transcription (post-processing). Prompt words influence the transcription engine itself, biasing it to recognize specific vocabulary during the acoustic decoding phase.