Skip to content
Projects

glimpse-hq/Glimpse has 64 GitHub stars and 6 forks. GitHot measured 0 stars in 24 hours and +2 in 7 days. Last checked 2026-11-15 22:26 UTC. Last push 2026-11-03.

How these counts are measured

What you can do

  • Local transcription. Turn your wifi off, it still works. ANE supported on Apple Silicon. Pick a model in Settings → Models.
  • Custom dictionary. Teach it names, brands, or terms.
  • Auto Dictionary. It picks up your custom words on its own.
  • Replacements. Say "my address," get 221B Baker Street.

Measured star growth

in 24h
0
in 7 days
+2
in 15 days
Update delayed
in 1 month
Update delayed

Net star changes over overlapping periods, not daily totals. Missing measurements are left blank. Stars measure attention, not software quality.

About

Voice dictation that stays on your device. Open-source, local-first, and built for people who actually care about privacy.

Productivity

README.md

Original

Glimpse

Voice dictation and meeting transcription that run on your computer.
Free and unlimited dictation, on Mac and Windows.

Glimpse

Download · Website · Pricing · Privacy

macOS 14+ Windows 10+, on the Microsoft Store


Hey! Glimpse is made mostly by me. I talk a lot faster than I type, and I wanted dictation that worked everywhere without sending my voice to someone's server or charging me every month. Nothing I tried really did all of that, so I built it.

Press your shortcut, talk, and your words show up wherever you're typing. It works offline, and dictation is free with no word limits on every model. I don't think you should have to pay to talk to your own computer. A license adds the extras, like recording meetings, AI cleanup and transcribing audio and video, and it's how I keep working on Glimpse.

Something broken or bugging you? Open an issue or email me at [email protected]. I do my best to respond ASAP.

Install

Screenshots

Glimpse home screen showing recent transcriptions Glimpse dictionary screen

Glimpse personalization screen Glimpse library screen for imported audio and video files

Features

Free, always

  • Local transcription. Turn your wifi off, it still works. ANE supported on Apple Silicon. Pick a model in Settings → Models.
  • Custom dictionary. Teach it names, brands, or terms.
  • Auto Dictionary. It picks up your custom words on its own.
  • Replacements. Say "my address," get 221B Baker Street.
  • History and search. Find anything you've dictated.
  • Import. Bring your dictionary, replacements, and history over from another dictation app.

With a license (14-day trial included)

  • Recording Mode. Record meetings, calls and lectures from your microphone and your computer's audio, or only the apps you pick, like Zoom or your browser. No bot joins the call. Follow a live transcript while you record, and drop bookmarks. More
  • Speaker detection. NVIDIA Nemotron 3 Diarization labels up to 8 speakers on your computer, live and after the recording. Runs on Metal, Vulkan or CPU, no NVIDIA card needed.
  • Library. Drop in audio or video, scrub the synced transcript, assign speakers, export to .txt, .md, .srt, or .vtt.
  • AI Cleanup. Polish dictated text with Glimpse's on-device cleanup model, or your own LLM in Settings → Providers.
  • Edit Mode. Highlight text, say what you want, and watch it rewrite in place.
  • Personalization. Different tones per app or site, with snippets for dynamic context.

License only (not included in the trial)

  • Local API. An OpenAI-compatible speech endpoint, running on your machine.
  • CLI. An optional glimpse command for the terminal.

Integrations

All integrations require a license.

Install Glimpse Raycast Extension Connect Glimpse MCP server

  • Your own. The CLI guide covers scripting Glimpse from Shortcuts, Finder, or anything else that can run a command.

Pricing

Edition Price For
Solo $25 one-time You, on 1 device
Plus $39 one-time You, on up to 3 devices
Business $60 / seat / year Work use, one seat per person, one device

Start with the 14-day trial, then buy or paste a license key in Settings → Account.

Privacy

Transcription stays on-device by default. Enabling an external speech or LLM provider sends audio or text directly to that provider. Your API keys stay local.

The app sends anonymous usage telemetry to PostHog EU to help prioritize development. It's tied to a random install ID, not your identity, and stored in the EU. Opt out anytime in Settings → App.

Exactly what's sent, and what isn't
  • Collected: app version and platform, launches and uptime, whether the app quit normally, durations and counts, which built-in features you use, whether microphone and accessibility permissions are granted, your dictation and display language settings, a coarse hardware class (memory size range, chip family, graphics vendor), country, and bounded error/crash categories. A crash also records a code location (source file and line, or module and offset) so we can find the bug. If the app crashed, the next launch reads the crash report your OS saved and sends only the error type and module and offset frames.
  • Never sent: transcripts, audio, API keys, prompts, raw error text or stacks, full file paths, microphone names, the apps you record, model or provider names you type in, provider endpoints, your IP address, or anything personally identifiable.

Opting out sends one final ping, then nothing, ever.

For the full picture, see the analytics wiki or analytics/. The website is a separate system with its own privacy policy.

Contributing

The Contributing Guide covers everything from translations to code to bug reports.

Questions, bugs, or feedback: [email protected] or GitHub Issues.

Acknowledgments

Speech models are downloaded in-app from Hugging Face. The live list lives in Settings → Models. By family:

The on-device cleanup model:

  • Glimpse C1 2B, Glimpse's adapter (Glimpse C1 License: free to use through Glimpse) on Ternary Bonsai 2 by PrismML (Apache-2.0), which is built from Qwen3.5 by Alibaba Cloud (Apache-2.0). Its LICENSE and NOTICE.txt are installed with the model files.

License

Glimpse is open source under the GNU AGPL-3.0. In short:

  • Use it freely. Run, study, modify, and redistribute it, including commercially.
  • Share your changes. If you distribute a modified version, or run one as a network service, its source must be available under AGPL-3.0.
  • Keep the attribution. Copyright and license notices stay intact, and a fork must state that it is based on Glimpse and link back to this repository.
  • Pick your own name. The Glimpse name and logo are trademarks and not covered by the license. Forks and redistributions need a different name and icon.
  • Bring your own Satoshi. The two Satoshi font files are not covered by the AGPL. They're in this repo under the ITF Free Font License with permission from Indian Type Foundry, and that permission is only for Glimpse.

The full license text and NOTICE are what apply; this summary is not a substitute for them.