Dictly

Product Introduction

1. Dictly is a privacy-focused dictation tool that converts speech into formatted text entirely on-device, eliminating cloud dependency while maintaining sub-100ms latency for real-time transcription. It operates offline across macOS, iOS, and iPadOS, using local processing for speech recognition, formatting, and storage.
2. The product prioritizes data sovereignty, workflow efficiency, and customization, enabling users to dictate securely in environments requiring strict privacy, such as healthcare, legal work, or air-gapped networks. Its core value lies in combining professional-grade formatting with zero data transmission, ensuring compliance with stringent security standards.

Main Features

1. Dictly processes speech 100% on-device using Apple’s Neural Engine, achieving ~100ms first-word latency and consuming less than 50MB RAM during active dictation, with no network usage even in airplane mode.
2. The Quick Capture overlay (invoked via ⌃⌘D on macOS) allows instant dictation from any app, with real-time text formatting through customizable Workflows like filler-word removal, Markdown conversion, or structured bullet lists.
3. Per-App Profiles automatically apply context-specific rules (e.g., code comments in VS Code, formal paragraphs in Mail), while custom dictionaries ensure accurate transcription of technical terms or names without cloud-based training.

Problems Solved

1. Dictly eliminates privacy risks associated with cloud-based dictation tools by ensuring voice data never leaves the device, addressing compliance needs for industries like healthcare (HIPAA) or legal (client confidentiality).
2. It serves professionals requiring rapid, structured text output—writers, developers, medical practitioners—who need offline-ready tools with minimal latency and maximal formatting control.
3. Use cases include drafting emails during flights, adding code comments without switching windows, or documenting patient notes in secure hospital networks where internet access is restricted.

Unique Advantages

1. Unlike cloud-dependent tools (e.g., Otter.ai, Dragon), Dictly’s offline architecture guarantees zero data leakage, with all processing handled by on-device ML models optimized for Apple Silicon.
2. The live transformation pipeline allows users to program cascading formatting rules (e.g., “remove filler words → apply title case → insert Markdown headings”) that execute in real time as they speak.
3. Competitive advantages include a 5MB install footprint, Swift-native performance, and configurable insertion modes (Type-out, Paste, Auto-Submit) that reduce post-dictation editing by 70% compared to manual formatting.

Frequently Asked Questions (FAQ)

1. Does Dictly require an internet connection? Dictly operates fully offline with identical functionality in airplane mode, as verified by macOS Network Inspector, and never transmits data to external servers.
2. Which devices and OS versions are supported? Dictly requires macOS 26+, iOS 26+, or iPadOS 26+, optimized for Apple Silicon chips (M1/M2/M3) to achieve sub-100ms latency during live transcription.
3. Can I customize formatting rules for specific apps? Yes, Per-App Profiles in Dictly Pro let you assign unique Workflows, dictionaries, and insertion behaviors (e.g., Auto-Submit in Slack) to individual applications like Notes or VS Code.
4. How is user data stored and secured? All transcriptions and settings are stored locally in SwiftData, with optional end-to-end encrypted CloudKit sync (roadmap feature) for cross-device profile sharing.
5. What languages does Dictly support? The current version supports English, with future language packs planned, all processed on-device using Apple’s speech recognition frameworks.