---
title: "v2026.6.18"
summary: "Added chunked streaming dictation for local STT backends so longer recordings are transcribed in bounded windows instead of retaining one large capture until push-to-talk release."
description: "Dictate v2026.6.18 release notes: what changed, and who contributed."
keywords:
  - Dictate
  - v2026.6.18
  - release notes
---

# v2026.6.18

<!-- Written by scripts/release/changelog.mjs from CHANGELOG/2026.6.18.md. Do not edit it: the release cut writes it again. -->

AI agents and tools can read these release notes as [plain Markdown](https://raw.githubusercontent.com/arcforgelabs/dictate/master/CHANGELOG/2026.6.18.md).

This release covers **1 pull request.**

## Highlights

- Added chunked streaming dictation for local STT backends so longer recordings are transcribed in bounded windows instead of retaining one large capture until push-to-talk release.
- Added live transcript events in the Settings UI, including partial/final transcript display and stale transcript suppression.
- Bounded audio/transcription queues and terminal recording caches to prevent long-running dictation sessions from growing memory without limit.
- Preserved stop-time audio callbacks and short final tails so streamed recordings do not lose the last spoken words.
- Kept hosted STT providers on final-audio transcription unless their backend explicitly opts into chunk streaming.

## Changes

- Added chunked streaming dictation for local STT backends so longer recordings are transcribed in bounded windows instead of retaining one large capture until push-to-talk release.
- Added live transcript events in the Settings UI, including partial/final transcript display and stale transcript suppression.

## Fixes

- Bounded audio/transcription queues and terminal recording caches to prevent long-running dictation sessions from growing memory without limit.
- Preserved stop-time audio callbacks and short final tails so streamed recordings do not lose the last spoken words.
- Kept hosted STT providers on final-audio transcription unless their backend explicitly opts into chunk streaming.

## Contribution record

[CHANGELOG/records/2026.6.18.md](https://github.com/arcforgelabs/dictate/blob/master/CHANGELOG/records/2026.6.18.md) lists every pull request in this release once, with its related issues and thanks.