Choose Handy if you want free local dictation whose code you can read, under the MIT licence, from a project that says it has no tracking telemetry. Choose Spokenly if you want free local dictation plus more options around it: a Local Only Mode switch that blocks network access, your own OpenAI, Deepgram or Groq keys at no charge from Spokenly, an iPhone app, and a Pro subscription for managed cloud models and sync. Both are free for local use with Whisper and Parakeet models, as of 3 October 2026, so price does not decide this one. The difference is how each earns your trust (public code against a documented switch) and how far each goes beyond plain dictation.
Publisher note: SuprFlow publishes this guide and is one of the products discussed. Competitor details come from each company's own pages, checked on 3 October 2026. Prices and features change; confirm before you buy.
We have not run a controlled benchmark of either app, so nothing here says which is more accurate.
Who should pick which?
- You want to read or build the code: Handy. Spokenly says it "is not open source and does not have a public GitHub repository".
- You want one switch that blocks the app's network access: Spokenly's Local Only Mode.
- You want cloud transcription with your own API key and no extra fee: Spokenly. Handy's documented cloud use is limited to optional post-processing of text.
- You also use an iPhone: Spokenly lists iOS. Handy lists Windows, macOS and Linux.
- You want managed cloud models and AI cleanup without keys: Spokenly Pro.
- You have an Intel Mac: Handy's latest release has an Intel Mac build. Check Spokenly's download page for yours.
How do Handy and Spokenly compare?
As of 3 October 2026. Handy details are from its README, releases page, site and privacy policy. Spokenly details are from its pricing page, Local Only Mode documentation, privacy policy and open-source statement.
| Handy | Spokenly | |
|---|---|---|
| Free local dictation | Yes: Whisper (Small, Medium, Turbo, Large) and Parakeet V3 | Yes: Whisper and Parakeet, "$0 forever", "Unlimited use", no account needed |
| Where speech is processed | On your computer: "without sending any information to the cloud" | On your device with local models; with cloud models, audio goes to Spokenly's backend, which "immediately forwards" it to third-party transcription services without storing it |
| Cloud options | Optional post-processing with a provider you choose; it receives the transcript | Your own keys (OpenAI, Deepgram, Groq and others) free of Spokenly charges, or Pro's managed models |
| Paid tier | None | Pro: $99.99 a year (shown as $8.33 a month) |
| What Pro adds | Not applicable | High-accuracy cloud models, sync across devices, AI text cleanup and Agent Mode, priority support |
| AI cleanup without paying Spokenly | Optional post-processing with a cloud provider you choose, Apple Intelligence, or a local endpoint, per its settings source | Apple Intelligence on the free local plan, or your own LLM key on the free keys plan, per its pricing table |
| Platforms | Windows, macOS, Linux. v0.9.8 (3 October 2026) has Apple Silicon and Intel Mac builds | macOS, iOS, Windows, Linux |
| Source code | MIT licence on GitHub | Proprietary |
| Analytics | "Does not contain advertising, behavioral analytics, or user-tracking telemetry" | Basic usage data such as button clicks and page views, never audio or transcripts; Local Only Mode blocks it |
| Network switch | None described; network use is update checks, model downloads and optional post-processing | Local Only Mode: blocks everything except model downloads and localhost |
| Refund | Not applicable | 14-day refund on Pro |
How do they earn trust differently?
Handy lets you check. Its code is public under MIT, and its privacy policy, effective 18 September 2026, lists all of its network use: update checks to GitHub Releases, model downloads from Hugging Face, and cloud post-processing only if you turn it on. A developer can confirm this in the source; a download is only proven to match the source if you build it. Is Handy safe? shows how to verify a release.
Spokenly gives you a switch. You cannot read its code, but its Local Only Mode disables "all network activity except local model downloads", allows localhost, and "blocks even these minimal analytics". If you pick an online model with it on, "transcription fails because network requests are blocked", so a wrong setting shows itself. Its privacy policy, updated 2 October 2026, says local-model audio and transcripts "never leave your device". Is Spokenly safe? goes further.
These are each project's own descriptions. We have not audited either app's network traffic.
What does each cost over three years?
List prices as of 3 October 2026, one person, USD, before tax.
| Option | Year 1 | After 3 years |
|---|---|---|
| Handy | $0 | $0 |
| Spokenly local models | $0 | $0 |
| Spokenly with your own keys | Provider usage fees | Provider usage fees |
| Spokenly Pro, yearly | $99.99 | $299.97 |
Spokenly's pricing page says it "charges nothing" for your own keys; the provider bills you. Handy's cloud post-processing works the same way. Spokenly pricing covers the student discount and billing options.
Choose Handy if…
- You want open-source code under a permissive licence.
- You want the fewest features and settings.
- You use an Intel Mac.
- You prefer a project that says it has no usage analytics at all.
Choose Spokenly if…
- You want a single switch that blocks network access.
- You want to plug in your own cloud transcription keys.
- You use an iPhone as well as a Mac.
- You may later want managed cloud models and sync, and accept a subscription for them.
How do you test both in 15 minutes?
Both are free. Use made-up content.
- Install both (4 minutes). Grant Microphone and Accessibility. Download the same kind of local model in each (for example, a Parakeet model) and note its name. Give the two apps different shortcuts.
- Keep it local. Turn on Local Only Mode in Spokenly. Leave Handy's post-processing off.
- Dictate in your mail app (3 minutes). "Hi Leocadia, the freight quote is 7,425 dollars, not 7,245. Delivery is Thursday the 16th, sorry, Friday the 17th. Copy Ms. Okonkwo-Haas."
- Check both names, both amounts and the corrected date.
- Turn Wi-Fi off and repeat (3 minutes). Both should keep working.
- Try your jargon (5 minutes). Add misses to each app's custom words or vocabulary and retry.
If they tie, decide on open source against the switch and the iPhone app.
A third option
Both apps already give free local dictation, so most readers do not need to pay anyone. SuprFlow is a paid, proprietary Mac app: hold Fn, speak, release, and text from an on-device model is pasted at the cursor, with Default formatting doing local cleanup of punctuation, capitalization and numbers. Optional cloud formatting sends transcript text, not audio, to the provider you pick. It costs $125 once for lifetime access on up to five Macs you own, with a 7-day refund on a first purchase and no trial. It is the wrong choice if you want free or open-source software, a network switch, an Intel Mac, or Windows, Linux or phone apps. It needs Apple Silicon and macOS 14 or later. SuprFlow vs Handy and SuprFlow vs Spokenly compare them.
Questions people ask
Are Handy and Spokenly really free?
Yes for local dictation. Handy is free and MIT licensed. Spokenly lists local Whisper and Parakeet models at "$0 forever" with no account. Spokenly Pro, for managed cloud features, is $99.99 a year as of 3 October 2026.
Is Handy or Spokenly more private?
Both keep audio on the Mac with local models. Handy's code is public and it says it has no tracking telemetry. Spokenly is proprietary but offers Local Only Mode, which blocks its basic analytics and any cloud model.
Is Spokenly open source like Handy?
No. Spokenly's site says it is not open source and has no public GitHub repository. Handy publishes its code under MIT.
Related guides
- Is Handy safe and open source?
- Handy alternatives
- Spokenly pricing
- Is Spokenly safe?
- Spokenly alternatives
- Best free dictation app for Mac