REAL

Holds up: Handy, Chatterbox and Recordly are real, free, open-source repos that cover the core jobs of Wispr Flow, ElevenLabs and Screen Studio, though each is a leaner tool than the paid product it is compared to

Facebook · Oct 2026

The claim'I found the opensource versions of: Wisprflow, Elevenlabs, Screen studio.' On screen: Handy is the open-source Wispr Flow and 'works on computer completely offline'; Chatterbox TTS by Resemble AI is the open-source ElevenLabs and 'also allows voice cloning'; Recordly is the open-source Screen Studio. Caption: 'needs much more attention! If you know more, put it in the comments!'

All three tools exist, are open source, and do roughly what the reel says. Handy (github.com/cjpais/Handy) is an MIT-licensed push-to-talk dictation app with about 33k stars, a v0.9.8 release on Oct 3, 2026, and installers for macOS, Windows and Linux. It runs Whisper or Parakeet models on your own machine, so the offline claim is true. What it does not copy is Wispr Flow's cloud AI cleanup that rewrites rambling speech into polished text; Handy gives you a transcript. Chatterbox (github.com/resemble-ai/chatterbox) is an MIT-licensed family of text-to-speech models from Resemble AI with about 27k stars, zero-shot voice cloning from a short reference clip, and a multilingual model covering 23+ languages. It is a model you run with Python and ideally a GPU, not a ready studio like ElevenLabs, and every clip carries Resemble's PerTh watermark. Recordly (github.com/webadderallorg/Recordly) is a desktop screen recorder with auto-zooms, cursor smoothing and styled frames for macOS 14+, Windows and Linux, about 33k stars, AGPL-3.0 per its README. Nothing is being sold here; the reel is an engagement post that asks for comments. The honest framing is 'free alternatives that cover the main job,' not 'the same product for free.'

What holds up

Reel read frame by frame: on-screen text names Handy ('speak into any text field, the free and open source app for speech to text'), Chatterbox TTS by Resemble AI (MIT badge), and Recordly (macOS, Windows, Linux, open source AGPL badge).
GitHub API, Oct 8, 2026: cjpais/Handy about 33.3k stars, MIT, latest release v0.9.8 (Oct 3) with .dmg, .exe, .msi, .deb, .rpm and AppImage builds; README says it 'works completely offline' using Whisper or Parakeet V3.
GitHub API and README: resemble-ai/chatterbox about 26.8k stars, MIT; Chatterbox-Turbo (350M, English), Nano (110M, CPU) and Multilingual V3 (500M, 23+ languages) support zero-shot voice cloning; built-in PerTh watermarking; install is pip install chatterbox-tts.
GitHub API and README: webadderallorg/Recordly about 32.9k stars, README badge says AGPL-3.0 (GitHub's license detector reports NOASSERTION); runs on macOS 14+, Windows 10 19041+ and Linux, with cursor hiding unsupported on Linux.

What doesn’t

Handy matches Wispr Flow's dictation, not its AI rewriting and formatting layer; expect raw transcripts.
Chatterbox is a model, not an app: you need Python, ideally an NVIDIA GPU, and you should only clone voices you have consent to use.
Recordly is AGPL-3.0, which matters if you plan to modify it and ship it inside a commercial product.

The catch

These are genuine free tools with large, active communities, but each one trades polish for price. The paid products still win on setup time, AI cleanup, hosted voices and support, so test the free version on your real workflow before cancelling a subscription.

How to actually do it

Dictation: install Handy from github.com/cjpais/Handy/releases (or brew install --cask handy, or winget install cjpais.Handy), pick Parakeet V3 for CPU machines or a Whisper model if you have a GPU, and set your shortcut.
Voice: try Chatterbox in the browser first at huggingface.co/spaces/ResembleAI/Chatterbox-Multilingual-TTS; if it fits, run pip install chatterbox-tts on a GPU machine and clone only your own voice or one with written consent.
Screen demos: download Recordly from its GitHub releases or recordly.dev, record a two-minute walkthrough, and compare export quality against Screen Studio before switching.

Every repo named in the reel was checked on GitHub on Oct 8, 2026: Handy and Chatterbox are MIT, Recordly is AGPL-3.0, and all three are actively maintained with roughly 27k to 33k stars. The comparison to paid products is fair for the core job and generous on polish.

Confidence: High
Posted by: a software-engineering creator on Facebook (reel showed about 141k views, 895 reactions); reel sent to Buddy Mon Oct 5, 2026, ~7:24 AM

See the original claim →

We test hype for free. We build the real thing for a living.

Thirty minutes, no pitch — and you'll leave with something useful either way.

Book a call with Todd or start with the free Business Checkup →

The Verdict Weekly

Three verdicts every Friday. Free forever, unsubscribe anytime, no spam — that would be ironic.