LumeVoice vs Spokenly: Free Local Dictation or Cleanup by Default?

LumeVoice vs Spokenly is mostly a question of where your speech turns into text. Spokenly can do it on your own computer or iPhone with free local models, or in the cloud through its Pro plan or your own API keys. LumeVoice always does it in the cloud, and puts its effort into what comes back: filler words, stutters and self-corrections are handled before the text lands in your app.

A quick disclosure: LumeVoice is our product, and we haven't used Spokenly ourselves, so nothing below is a hands-on test. What we say about Spokenly comes from its website, docs, policies, release notes and store listings, all checked on October 20, 2026. Spokenly changes quickly (its Mac app had nine releases between October 1 and October 18 alone), so confirm details on Spokenly's pricing page before you buy.

Who should pick which#

Pick Spokenly if you:

• want to dictate as much as you like for free, using local Whisper and Parakeet models with no word cap
• need to dictate without an internet connection on a Mac, Windows PC, Linux machine or iPhone
• use Linux, or want a Chrome extension
• already pay an AI provider such as OpenAI, Deepgram or Groq and want to use your own API keys
• want power-user extras on a Mac: voice answers for coding agents over MCP, a command-line tool, Agentic Actions and meeting recording

Pick LumeVoice if you:

• use an Android phone; Spokenly has no Android app yet
• want filler words, stutters and self-corrections ("five, actually six" becomes "six") cleaned up without choosing models or writing AI instructions
• want a one-time license that includes cloud dictation
• are setting up an older Mac: LumeVoice runs on macOS 11.5 or later, while Spokenly needs macOS 13.3
• find typing hard because of a disability, injury or health condition: LumeVoice gives 50% off, with no proof needed

LumeVoice vs Spokenly at a glance#

Spokenly details as of October 2026, from spokenly.app, its docs and its store listings.

Price and 3-year cost#

Spokenly's pricing comes in three layers (Spokenly pricing):

• Free costs nothing and needs no account: local Whisper and Parakeet models with no word cap, plus your own API keys from providers such as OpenAI, Deepgram and Groq, with no Spokenly fee on top (the provider bills you).
• Local+ is a one-time $29 for the Notch and Capsule recording styles, settings sync and priority support, without Spokenly's cloud models.
• Pro is $9.99 a month or $99.99 a year on Spokenly's website. It adds Spokenly-managed cloud models such as GPT-4o Transcribe, Deepgram Nova 3 and Soniox, hosted AI cleanup and everything in Local+.

In the iPhone app, the US App Store lists Pro at $9.99 or $10.99 a month and $49.99 or $54.99 for six months (App Store listing).

What each option costs if you keep it for three years:

If Spokenly's local models are good enough for you, nothing here beats it. The fairer comparison with LumeVoice is Spokenly Pro, the plan where Spokenly does the cloud transcription and cleanup for you: it's cheaper month to month ($9.99 against $14.99), and on yearly billing it's $99.99 against $79.99/year. Only LumeVoice sells a one-time license that covers its cloud dictation.

Discounts. Spokenly gives verified students 50% off Pro (student discount) and teams of two or more 30% off every seat for as long as they subscribe (team plan). Its site doesn't list a disability discount.

LumeVoice gives students and teachers 50% off, and gives 50% off to anyone who finds typing hard because of a disability, injury or health condition, with no proof needed. Codes are entered at checkout on Mac or Windows. See the accessibility discount, student pricing and LumeVoice pricing.

Platforms: where each app runs#

Mac. LumeVoice lives in the menu bar (hold fn, talk, let go) and runs on macOS 11.5 or later. Spokenly needs macOS 13.3 or later and listens for Right Command by default (Spokenly for Mac). Get Spokenly from its website: it says its Mac App Store version can't paste automatically and is no longer updated (Spokenly docs). More on LumeVoice for Mac.

Windows. Both support Windows 10 and 11. LumeVoice sits in the system tray: hold Ctrl + Win, tap Space while holding it for hands-free, or press Esc to cancel, and it installs without admin rights. Spokenly installs from the Microsoft Store or a direct installer, and says its Mac and iPhone apps are its most mature, with the Windows build covering the core dictation and transcription features (Spokenly for Windows). See LumeVoice for Windows.

Linux. Only Spokenly, with deb and rpm packages for x86_64 and ARM64, an x86_64 AppImage, and support for X11 and Wayland (Spokenly for Linux).

iPhone and iPad. Both work as keyboards. The Lume keyboard needs iOS 17 or later. Spokenly's runs on iPhone and iPad from iOS 16.4, per its App Store listing, and can use a local Parakeet model without a connection (Spokenly for iPhone). See LumeVoice for iPhone.

Android. Only LumeVoice, as a floating bubble over the keyboard you already use. Spokenly's Android page offers a waitlist (Spokenly for Android). See LumeVoice for Android.

In the browser. Spokenly has a Chrome extension, with 4,000 users and 4.4 stars from 41 ratings on the Chrome Web Store as of October 2026; its free mode uses Chrome's own speech recognition (Chrome Web Store). LumeVoice has no extension; on a Mac or PC, its desktop app types into any website instead.

Features: cleanup, models and extras#

Both apps type where your cursor is, add punctuation, keep a history, transcribe audio and video files on the desktop, and let you fix words they keep getting wrong. The differences:

Cleanup. LumeVoice cleans up every dictation by default: it removes filler words and stutters, applies self-corrections ("let's meet at five, actually six" becomes "Let's meet at six."), turns spoken lists into lists, and fixes grammar and spelling. Spokenly leaves more of this to you. Its modes run AI Instructions you write, through its hosted AI on Pro, your own AI provider or Apple Intelligence (Modes). Its new Local Model Cleanup removes fillers and applies spoken corrections on your computer, but it's off by default, works best in English and on a Mac needs Apple silicon and macOS 15 (Local Model Cleanup).

Models and keys. This is where Spokenly shines: you can pick a model per mode, from local Whisper, NVIDIA Parakeet and Apple's Speech Analyzer to cloud models from OpenAI, Deepgram, Soniox, ElevenLabs and others, with your own keys from 11 providers. LumeVoice makes that choice for you: Whisper large-v3-turbo running on Groq, with Mistral's Voxtral as a fallback, and a language model for the cleanup.

Power-user extras. Mostly on the Mac, Spokenly adds Agentic Actions (a mode can search the web, open apps or run a shell command after you speak), an MCP server that lets Claude Code, Codex and Cursor ask you questions you answer by voice, a command-line tool, and meeting recording without a bot (Agentic Actions, voice for agents). LumeVoice doesn't control your computer or record meetings; it writes down what you said.

Your words. LumeVoice's custom dictionary feeds names and jargon to speech recognition on Mac and Windows, and snippets type saved text, such as an address, when you say a short cue. Spokenly has Word Replacements, which can use regular expressions, plus keyterms for supported cloud models (Word Replacements).

Languages. LumeVoice works in 100 languages, detects them automatically and can give you text in a different language from the one you spoke. Spokenly lists 100+ languages and says quality varies by model and language; translation is something you set up as a mode.

Privacy: where your voice goes#

With Spokenly, privacy depends on the path you choose (privacy policy, updated October 18, 2026):

• Local models: your audio and transcripts stay on your device. If a mode uses AI Instructions, the transcript can still be sent to an AI service.
• Spokenly's cloud models (Pro): your audio goes to Spokenly's backend, which doesn't store it and passes it to a third-party transcription provider.
• Your own API keys: your audio goes straight from your device to the provider you picked, under that provider's terms.

Spokenly's analytics never include audio or transcripts. Local Only Mode blocks network requests (except local model downloads), analytics included, and organizations can lock it on with MDM, Group Policy or Intune (Local Only Mode). Note that Spokenly's own summary says it uses cloud models by default, so download a local model before you count on it offline (Spokenly's summary).

LumeVoice is cloud-based. Your audio is sent for processing and isn't stored afterwards, nothing you say is used to train models, and your dictation history stays on your device. It needs an internet connection.

If your audio must stay on your machine, Spokenly with a local model is the right pick. LumeVoice has nothing like it.

Where Spokenly is the better choice#

• Free, unlimited dictation with local models, and no account.
• Offline use on Mac, Windows, Linux and iPhone, with a Local Only Mode that companies can enforce.
• Linux and a Chrome extension.
• Your own API keys with no Spokenly fee, and a long list of models to choose from.
• Mac power features: Agentic Actions, MCP voice input for coding agents, a command-line tool and meeting recording.
• A $29 one-time option if the free tier plus a few extras is all you need.
• Team seats at 30% off.

Where LumeVoice is the better choice#

• Android, which Spokenly doesn't support yet.
• Cleanup without setup: filler words, stutters and self-corrections handled on every dictation, in 100 languages.
• A one-time license that includes cloud dictation at $249. Spokenly's one-time Local+ leaves out its cloud models.
• Older Macs: macOS 11.5 or later, where Spokenly needs macOS 13.3.
• An accessibility discount of 50% off, with no proof needed; Spokenly doesn't list one.
• Fewer decisions: no models, keys or provider bills to manage.

Switching from Spokenly to LumeVoice#

1. Install LumeVoice next to Spokenly on the computer and phone you use most.
2. Give each app its own shortcut. On a Mac they don't clash by default (Spokenly listens for Right Command, LumeVoice for fn). On Windows, Spokenly's Microsoft Store listing shows Ctrl + Win, which is also LumeVoice's default, so change one of them in Settings.
3. Bring your words across. Add the names and terms from Spokenly's Word Replacements to LumeVoice's custom dictionary, and turn phrases you repeat into snippets.
4. Dictate real work for a few days using your free words.
5. Ask for a discount code before you pay if you're a student or find typing hard, then upgrade on Mac or Windows.
6. Cancel Spokenly Pro if you had it. Website purchases can be refunded within 14 days, and App Store subscriptions are cancelled through Apple (refund policy). If you only used local models, there's nothing to cancel.

For the wider field, see our Spokenly alternatives, the Spokenly pricing breakdown and Spokenly vs Wispr Flow. Every other head-to-head is on our comparisons page.

The bottom line#

Spokenly is the more flexible app and, for many people, the cheaper one. Its local models cost nothing, work without a connection and keep your audio on your device, and it reaches Linux and Chrome.

LumeVoice is narrower: cloud only, no Linux, no extension. It's built so that what lands on the page is already clean, and it covers Android.

Choose Spokenly if free use, offline dictation or Linux is what you need.

If you'd rather not configure anything, or your phone runs Android, give LumeVoice a week of real writing and keep whichever app leaves you less to fix.