Free Voice-to-Text Engine & Documentation Automator

Dictate.
Transcribe.
Transform.

More than just dictation. Dicty is a complete voice engine that turns your spoken words, audio files, and videos into perfectly formatted text—for free. Effortlessly build and maintain living documentation by attaching your voice directly to your codebase, Git PRs, and tickets.

dicty · new capture
Recording00:42

“We're moving the API observer to an event-driven queue so retries don't block the main thread…”

Attached artifacts

PR #142: Refactor API observer
commit a3f9c1 · queue adapter
LIN-284 · Event-driven retries
architecture/api-observer.mdsynced
# API Observer → Event Queue

Migrated the observer to an event-driven queue to decouple retries from the request path.

## Rationale
  • — Retries no longer block the main thread
  • — Backpressure handled by the adapter
Refs:
→ PR #142 · a3f9c1 · LIN-284

The Architecture

Built for how engineers actually think.

Five core systems that turn scattered spoken intent into durable, connected knowledge, or handle your everyday transcriptions flawlessly.

Zero-Friction Voice Capture

Push hotkey, speak your thoughts, done. Captures the 'Why' behind code changes without losing flow or breaking context.

Traceable Context Selector

Ground your docs in reality. Select related PRs, commits, and branches so every doc links directly to production code.

Hub-and-Spoke Orchestration

Prevents documentation decay. Reconciles new updates with existing architecture files rather than creating redundant, linear logs.

Storage Agnostic Sync

Your docs, your vault. Seamless native export to Obsidian, Notion, Confluence, or GitHub Wikis.

General-Purpose Voice Engine

Beyond coding. Import audio/video files for transcription, translate between languages on the fly, and use local Ollama or Cloud LLMs to perfectly format your text—all for free.

The Architectural Librarian

Knowledge that compounds, instead of decaying.

Most tools append. Every note becomes another orphaned log nobody reads. Dicty reconciles new intent into a living hub of architecture docs — so the truth stays in one place.

Linear logging

Append-only. Redundant. Decaying.

2026-02-01 · Auth rework notes
2026-02-03 · Auth notes (v2)
2026-02-09 · Re: auth notes
2026-02-14 · Auth — actually final

Four documents describe the same system. Which one is true? Nobody knows.

Hub-and-spoke

Reconciled into one source of truth.

VKE Hub
PR #482
Standup
Commit a1f9
Incident
Voice note
ADR-014

Every spoke feeds the hub. One canonical doc, always current, always linked to code.

How it Works

From spoken word to synced doc in four steps.

Step 1

Speak

Trigger via local shortcut and dictate your architectural intent or task context.

Step 2

Select Artifacts

Attach relevant GitHub PRs, commits, or ticket links.

Step 3

Reconcile

Dicty's AI synthesizes the diffs and voice input into structured markdown.

Step 4

Direct Sync

Push updated docs straight to your target knowledge storage.

Integrations

Works seamlessly with your tech stack and documentation vaults.

GitHub
GitLab
Notion
Obsidian
Confluence
Linear
Jira
Slack

Pricing

Simple, transparent pricing. No forced accounts.

Starter

Free Trial

$0/ 14 days
  • 14 days full access
  • Local transcription engine
  • Basic MDX export
Download Now
Most popular

Pro

$4/ month
  • Unlimited Voice-to-Knowledge generations
  • Full Git PR / Issue Tracker context integration
  • Multi-vault sync (Notion, Obsidian, Confluence, GitHub)
Temporarily Unavailable

Early Supporter Lifetime

Limited time pricing

$99one-time
  • All Pro features + priority updates
  • Data sovereignty (100% local processing)
  • No recurring fees
Temporarily Unavailable

Looking to equip your entire engineering team? Contact us for Team Licensing