Free and open source, like TypeWhisper. Ours needs no setup.
TypeWhisper's core app is GPLv3, the same license we use, and its free tier is genuinely generous: seven speech engines, Windows and Mac, and a build-your-own cleanup system called Workflows. EnviousWispr offers two engines, switched manually in Settings, and ships EG-1, a model built and measured specifically for dictation, with nothing to configure.
Quick answer
TypeWhisper or EnviousWispr: which should you choose?
Both apps have free GPLv3 cores and can run Mac dictation locally. Choose TypeWhisper if you want seven speech engines, Windows support, file transcription, or Workflows that connect your own prompts, providers, translation, and automation steps. Choose EnviousWispr if you want two curated speech engines plus a purpose-built polish model that works without building a workflow. TypeWhisper is the stronger toolkit for people who want to configure the pipeline. EnviousWispr is the simpler option for people who want to install the app, choose a hotkey, and get polished text at the cursor with fewer decisions. The meaningful difference is control: TypeWhisper exposes more of the system, while EnviousWispr makes more choices for you. Both can keep ordinary dictation on the Mac.
Still deciding? See the best Mac dictation apps by use case or browse all 17 dictation app comparisons.
Feature comparison
An honest look at how the two tools stack up across the dimensions that matter most.
| EnviousWispr | TypeWhisper | |
|---|---|---|
| Price (core app) | Free. No paid tier at all. | Free (GPLv3). Paid Commercial license from EUR 5/mo unlocks non-GPL terms, Cloud Folder Sync, and Automatic Correction Learning. |
| Account required | No | No |
| Source code | Open source (GPLv3) | Open source (GPLv3, core) |
| Speech engines | Parakeet TDT v3 + WhisperKit, switch in Settings | Seven options: WhisperKit, Parakeet TDT, Voxtral, Qwen3 ASR, IBM Granite Speech, Apple Speech, plus ONNX-based engines on Windows |
| AI cleanup (after transcription) | EG-1, purpose-built and measured, ships ready | Workflows: configurable, sends your transcript through an LLM you set up (OpenAI, Groq referenced) |
| Needs a prompt to be written | No | No for common tasks: the installed build includes presets for lists, email, meeting notes, translation, JSON, and more; custom Workflows remain configurable |
| Learns your corrections automatically | No | Yes, Dictionary Autocorrection, Commercial license only |
| Live preview while speaking | Yes, off by default | Yes, live preview engine override |
| Recover a cancelled dictation | Yes (Escape Recovery, opt-in, keybind/triple-press cancel only) | Not documented |
| Custom vocabulary | Yes (imports from 8 other apps, TypeWhisper included) | Yes, with per-word case sensitivity and enable/disable state |
| Transcription latency | 0.61s median; ~1.65s with on-device AI polish | Not published |
| Platform support | macOS (Apple Silicon only) | macOS and Windows, dedicated builds for each |
| System-audio recorder | No | Yes, microphone, system audio, or both, with mixed or separate tracks |
| Watched-folder transcription | No | Yes, automatic transcription and text or subtitle export |
| Meeting automation | No | Yes, calendar reminders or automatic recording when a matching meeting starts |
| Dictionary and snippet sync | No | Yes, with a Commercial license, through a private iCloud container or a user-chosen synced folder |
TypeWhisper claims sourced from typewhisper.com, its pricing page, its changelog, the GitHub API, and the installed TypeWhisper 1.6.0 bundle. The live window could not be captured, so exact UI labels were not firsthand-tested. EnviousWispr latency from production PostHog data on Apple Silicon Macs. Competitor claims last verified: 2026-08-21.
Both are free and GPLv3. Here is what EnviousWispr adds
TypeWhisper's free tier is unusually generous. This is what separates the two apps beyond that.
EG-1 is fine-tuned specifically for dictation cleanup and runs on the same hotkey as transcription. TypeWhisper's Workflows are more flexible, and the installed build includes ready-made actions for common writing tasks alongside custom workflows.
EG-1 builds a real bulleted list out of a spoken list 83% of the time, up from under 1% before it was tuned, and correctly keeps only what you meant when you self-correct mid-sentence 78% of the time. We found no equivalent published benchmark for a TypeWhisper Workflow, since results depend on which provider and prompt you set up.
Escape Recovery keeps a dictation you cancel with your keybind or a triple press, instead of discarding it instantly. Off by default; you choose to turn it on. We found no equivalent documented for TypeWhisper.
TypeWhisper's GPLv3 core is free, but Cloud Folder Sync, Automatic Correction Learning, and non-GPL terms sit behind a Commercial license starting around EUR 5/mo. EnviousWispr has no paid tier of any kind.
Built-in filler-word removal runs before AI polish, and it is language-aware: with your language locked to German, Dutch, Danish, or Norwegian in Settings, it keeps real words like "er" and "um" instead of stripping them as English filler.
Smart Import reads your custom vocabulary straight out of TypeWhisper's own files, without changing them, and shows you every word before anything joins your library. No retyping.
Where does your voice go?
Both apps process your voice on your Mac by default. For a deeper dive, read on-device vs cloud dictation privacy.
Parakeet TDT: built for speed
EnviousWispr uses NVIDIA Parakeet TDT v3, a multilingual ASR model covering 25 European languages that runs natively on the Apple Neural Engine. No network dependency, no server queues.
Based on production data from Apple Silicon Macs. Results vary by hardware and settings. TypeWhisper does not publish comparable latency figures.
A toolkit, or a finished tool
TypeWhisper gives you the pieces to build the dictation setup you want. EnviousWispr hands you one that already works.
TypeWhisper offers seven engine options across Mac and Windows: WhisperKit, Parakeet TDT, Voxtral, Qwen3 ASR, IBM Granite Speech, Apple Speech, and ONNX-based engines. If you want to compare these engines directly, or you have a strong preference for one, that choice is real.
EnviousWispr runs two: Parakeet TDT v3 for the 25 languages it covers best, WhisperKit for the rest. There is no larger menu; you switch between the two in Settings, Transcription.
TypeWhisper's Workflows are a genuine automation system: cleanup, translation, extraction, and formatting steps, referencing providers like OpenAI and Groq in its own changelog. The installed build also includes presets for Action Items, Create Table, email drafting, lists, JSON, Meeting Notes, Reply, and Translate.
EG-1 is a single model, fine-tuned by us specifically for the problem of turning spoken dictation into clean text, shipped with the app, with results we measure and publish. It runs on the same hotkey as transcription with nothing to configure.
Choose TypeWhisper if any of this matters more to you
TypeWhisper's free tier is unusually generous, and its flexibility is real. It may be the better fit here:
TypeWhisper has a dedicated Windows build. EnviousWispr is macOS only.
Seven engines to pick from, including specialized ones like Qwen3 ASR and IBM Granite Speech. EnviousWispr offers two, switched manually in Settings.
TypeWhisper's Commercial license adds Dictionary Autocorrection, which learns from your confident manual edits and applies them going forward. We do not offer an equivalent.
Workflows let you wire up your own provider, prompt, and automation steps, including translation and app-specific behavior. If that flexibility is what you are after, EG-1's single tuned model will not give it to you.
If you want TypeWhisper's seven engine options, need Windows, or want the flexibility to build your own cleanup workflow, TypeWhisper is a strong pick. If you want a purpose-built polish model with a published, measured track record, give EnviousWispr a try. You can always use both.
Common questions
Yes. TypeWhisper's core app is free and open source under GPLv3, the same license we use, with core dictation, its speech engines, and its Workflows automation system all included at no cost. TypeWhisper also sells a Commercial license, starting around 5 EUR a month, that unlocks Cloud Folder Sync and Automatic Correction Learning and lets a business use it under non-GPL terms. EnviousWispr has no paid tier at all.
Both are GPLv3, free at the core, and run on-device. TypeWhisper offers seven engine options across Mac and Windows: WhisperKit, Parakeet TDT, Voxtral, Qwen3 ASR, IBM Granite Speech, Apple Speech, and ONNX-based engines, plus a Workflows system for building your own cleanup, translation, and formatting steps. EnviousWispr runs two engines, Parakeet TDT and WhisperKit, switched manually in Settings, and EG-1, a model built and measured specifically for cleaning up dictation, with no setup required. TypeWhisper also runs on Windows in addition to Mac; EnviousWispr is Mac only.
TypeWhisper's equivalent is called Workflows, a configurable system that can send your transcript through an LLM for cleanup, translation, or formatting. The installed 1.6.0 build also contains ready-made actions for lists, email, meeting notes, translation, JSON, and more. A separate Dictionary Autocorrection feature for paid Commercial licenses learns from confident edits. EG-1 is purpose-built for dictation cleanup and ships ready inside EnviousWispr.
Yes. After the one-time model download, transcription runs entirely on-device with no internet required. AI polish can also run offline using EG-1 (our own model, macOS 14+), Apple Intelligence (macOS 26+), or Ollama with a local model. Cloud providers like OpenAI, Claude, or Gemini are optional and require your own API key.
Yes, and under the same license. Both EnviousWispr and TypeWhisper's core app are licensed under the GNU General Public License v3 (GPLv3). TypeWhisper also offers a paid Commercial license for organizations that need non-GPL terms; EnviousWispr has no such tier and no non-GPL option.
Yes. EnviousWispr's Smart Import reads your custom vocabulary directly out of TypeWhisper, without changing TypeWhisper's own files, and shows you every word it found before anything joins your library. Entries you switched off in TypeWhisper are skipped rather than imported, and per-word case sensitivity comes across for the entries that do; TypeWhisper's match-strictness setting does not. It also imports from Wispr Flow, Superwhisper, FluidVoice, Handy, Spokenly, Juno, and Vox.
Yes, with its own dedicated Windows build, separate from the macOS app but sharing most of its features. EnviousWispr is macOS only, on Apple Silicon.
EnviousWispr runs NVIDIA Parakeet TDT v3 for 25 European languages and WhisperKit for 99+ languages, switched manually in Settings. TypeWhisper offers a much wider menu: WhisperKit, Parakeet TDT, Voxtral, Qwen3 ASR, IBM Granite Speech, and Apple Speech on Mac, plus ONNX-based engines on Windows, with cloud engines available through add-ons.
Yes. Download EnviousWispr, set your keybind, and Smart Import can bring your TypeWhisper custom words across in one pass. There is no other data to migrate for dictation purposes. Both apps work in any text field on macOS.
Compare with other tools
Free, GPLv3, and nothing to configure.
Free to download. No account required, no paid tier, no workflow to build.