OpenCluely v1.8.6

Installation

Choose the artifact for your platform:

Platform	File	Notes
Windows	OpenCluely-Setup-*.exe	NSIS installer — installs app + adds to Start Menu
Linux	*.deb	Debian/Ubuntu — auto-pulls system deps (Python, ffmpeg, GTK)
Linux	*.AppImage	Universal — no install, just chmod +x and run

macOS: no pre-built build is shipped. The app is unsigned/un-notarized, so macOS Gatekeeper blocks it as "damaged". Run OpenCluely from source instead:

git clone https://github.com/TechyCSR/OpenCluely && cd OpenCluely && ./setup.sh

Requires Node.js 18+. See the README for details.

First Run

On first launch, OpenCluely opens the Settings window automatically for your Google Gemini API key.
Get a free key from Google AI Studio. You can also edit .env directly.

Optional: Local Whisper (offline speech)

Install Python 3.10+ and ffmpeg, then run:

./setup.sh --skip-install-system-deps

Checksums

See SHA256SUMS.txt attached to this release.

What's Changed

Summary

This release includes 5 commits since v1.8.5.

Other Changes

• Warn that screen-capture invisibility is unsupported on Linux; document limitation (8f48258)
• Moved Limitaion sections under Troublesshooting (eba150d)
• Render LaTeX math in overlay responses (subscripts, superscripts, symbols) (43de620)
• Render LaTeX math in chat too via shared lib/mathrender.js (ffbe8f3)
• Ship Windows+Linux only, fix missing Windows artifact, direct macOS to build from source (dffdf1a)

Full Changelog: v1.8.5...v1.8.6

OpenCluely v1.8.5

Installation

Choose the artifact for your platform:

Platform	File	Notes
Windows	OpenCluely-Setup-*.exe	NSIS installer — installs app + adds to Start Menu
macOS (Apple Silicon)	*-arm64.dmg	M1/M2/M3/M4 Macs
macOS (Intel)	*-x64.dmg	Older Intel Macs
Linux	*.deb	Debian/Ubuntu — auto-pulls system deps (Python, ffmpeg, GTK)
Linux	*.AppImage	Universal — no install, just chmod +x and run

First Run

On first launch, OpenCluely opens the Settings window automatically for your Google Gemini API key.
Get a free key from Google AI Studio. You can also edit .env directly.

Optional: Local Whisper (offline speech)

Install Python 3.10+ and ffmpeg, then run:

./setup.sh --skip-install-system-deps

Checksums

See SHA256SUMS.txt attached to this release.

What's Changed

Summary

This release includes 2 commits since v1.8.1.

Fixes

• Fix broken SDK in Electron: use HTTPS SSE streaming and HTTPS-first requests (9fbb914)

Other Changes

• Speed up AI: Flash-Lite model, disable thinking, drop preflight, stream chat+screenshots (3516dc5)

Full Changelog: v1.8.1...v1.8.5

OpenCluely v1.8.1

Installation

Choose the artifact for your platform:

Platform	File	Notes
Windows	OpenCluely-Setup-*.exe	NSIS installer — installs app + adds to Start Menu
macOS (Apple Silicon)	*-arm64.dmg	M1/M2/M3/M4 Macs
macOS (Intel)	*-x64.dmg	Older Intel Macs
Linux	*.deb	Debian/Ubuntu — auto-pulls system deps (Python, ffmpeg, GTK)
Linux	*.AppImage	Universal — no install, just chmod +x and run

First Run

On first launch, OpenCluely opens the Settings window automatically for your Google Gemini API key.
Get a free key from Google AI Studio. You can also edit .env directly.

Optional: Local Whisper (offline speech)

Install Python 3.10+ and ffmpeg, then run:

./setup.sh --skip-install-system-deps

Checksums

See SHA256SUMS.txt attached to this release.

What's Changed

Summary

This release includes 1 commits since v1.8.0.

Other Changes

• Stop electron-builder auto-publishing so changelog release body is used (802de18)

Full Changelog: v1.8.0...v1.8.1

1.8.0

Installation

Choose the artifact for your platform:

Platform	File	Notes
Windows	OpenCluely-Setup-*.exe	NSIS installer — installs app + adds to Start Menu
macOS (Apple Silicon)	*-arm64.dmg	M1/M2/M3/M4 Macs
macOS (Intel)	*-x64.dmg	Older Intel Macs
Linux	*.deb	Debian/Ubuntu — auto-pulls system deps (Python, ffmpeg, GTK)
Linux	*.AppImage	Universal — no install, just chmod +x and run

First Run

On first launch, OpenCluely opens the Settings window automatically for your Google Gemini API key.
Get a free key from Google AI Studio. You can also edit .env directly.

Optional: Local Whisper (offline speech)

Install Python 3.10+ and ffmpeg, then run:

./setup.sh --skip-install-system-deps

Checksums

See SHA256SUMS.txt attached to this release.

What's Changed

Summary

This release includes 15 commits since v1.7.0.

Features

• Add monochrome brand favicon and social preview image with generator (78ce54c)
• Add star indicator to navbar GitHub pill and README star line (f0bf3b7)

Fixes

• Fix install step overflow that stretched webapp on small screens (4c63edd)
• Fix Whisper flush queue stranding utterances during in-flight transcription (699a4bb)
• Fix VAD noise-floor seeding so immediate speech never stalls detection (30512b3)
• Fix stranded streaming bubble when transcription LLM fully fails (1034434)
• Fix Quit menu API call and update stale Google AI Studio link (0c226c6)

Improvements

• Polish webapp UI: logo marquee, glassy footer, menu toggle, fixes (f41b0d1)

Other Changes

• Redesign webapp: minimal B&W, authentic overlay hero, dynamic downloads, SEO (62aa61f)
• Rewrite README in plain human prose without emojis (07d071d)
• Cancel Gemini stream after timeout to stop late chunks leaking to UI (c75bc6b)
• Let onboarding continue when Gemini key already set in .env (5723a45)
• Open external links in system browser instead of app windows (7ca8a06)
• Show full commit range in release notes, not just last commit (666cb22)
• Stop double-saving settings on speech provider change (acda7d7)

Full Changelog: v1.7.0...v1.8.0

1.7.0

v1.7.0

Update README: live download badges, mark completed TODO items

1.6.0

v1.6.0

Fix cross-platform speech bugs: Windows spaced-path probe, macOS mic …

1.5.5

Remove Gemini API test button and Whisper mic test

- Remove Test connection button from onboarding API key screen
- Remove Whisper 5-second mic test panel and state from model-download screen
- Remove unused test-whisper-recording IPC handler, preload bridge, and speech.service method
- Keep model-download choice (now/later) and actual turbo weight download intact
- Downloaded turbo model is used for live speech without requiring user verification

1.5.3

v1.5.3

Fix stray 'login' token in config.js causing app crash on startup

1.5.2

v1.5.2

Fix Gemini API test, add model download screen, fix mic button visibi…

1.5.1

v1.5.1

Fix install bugs: friendly Gemini errors, install button states, mic …