Show HN: BetterWispr – Free, Open-source Dictation For Mac

TL;DR

Before you orderOffer from Amazon

Get the latest gadgets delivered free with Prime

• Fast, free delivery on millions of items
• Prime Video, Amazon Music and more included
• Member-only deals all year

Start your free Prime trial →Free trial for eligible customers · Cancel anytime

As an affiliate, we earn on qualifying purchases.Hide

BetterWispr has been presented on Hacker News as a free, open-source dictation app for macOS 14 and later. It records speech through a keyboard shortcut, transcribes it with a selected model on the Mac, and can clean up text before inserting it into another app.

BetterWispr, a free, open-source dictation app for Mac, is being introduced through a Show HN launch post. The app lets users hold a keyboard shortcut to record speech, then transcribes and inserts the result into the active application using a selected speech model; its maker says the speech processing can stay on the Mac.

The app requires macOS 14 or later, with Apple Silicon recommended for running local models. Its default shortcut is Option + Space: users hold the keys while speaking and release them to finish recording. The product information says transcription starts after recording stops, rather than appearing word by word during speech.

BetterWispr offers nine built-in model options from Apple speech, Parakeet and Whisper, according to its website. Users can also configure supported external APIs, including services compatible with OpenAI’s API. The maker says local models are downloaded to the Mac and can then work offline, with no account or cloud transcription fallback required for local transcription.

Optional features include meeting notes, per-app writing styles, English filler-word cleanup, voice commands and a local history of original and corrected text. The site says meeting summaries can be generated using Apple Intelligence or Ollama on the Mac. Its license is Apache 2.0, and the project invites users to inspect the source code, report issues or build the app themselves.

A Local Option for Mac Dictation

BetterWispr’s main distinction is its combination of dictation, editable cleanup and local processing. Users who want speech-to-text without sending recordings to a transcription service may find the on-device approach relevant, while the open-source license offers a way to inspect how the software is built. These are product characteristics described by the maker, not independent evaluations of privacy or accuracy.

The app also gives users control over how closely the result follows their speech. The maker describes a “Light” setting that removes English filler words, a “Medium” setting that also edits for clarity using a chosen notes model, and a “None” setting that keeps every word. Styles can be configured for different apps, but the site says cleanup, styles and voice commands operate in English.

Those choices matter because dictation tools can change meaning when they edit spoken language. Keeping an original transcript in local history, and allowing users to disable cleanup, could help users check or restore what they said. The practical value will depend on transcription quality, language needs, hardware and how well text insertion works in each user’s apps.

How the Dictation Workflow Works

BetterWispr is designed to type into the application the user was already using. The maker says users place the cursor in a text field, hold the shortcut, speak and release; the app then transcribes the recording and inserts the text. If automatic pasting is blocked, the product says the transcript can be copied from its dashboard instead.

Some functions require macOS permissions. The maker lists microphone access for recording, Accessibility permission for inserting text into other apps, and Speech Recognition permission when using Apple speech. A compatible local model and required assets must be installed before offline transcription is available. The amount of storage used depends on the model selected.

Language coverage varies by model. The website says Parakeet TDT v3 supports 25 European languages, a separate Parakeet option supports Japanese, and Whisper offers broader multilingual recognition. The maker cautions that Apple speech availability depends on locale and installed system assets. Its site describes an interface demonstration as an illustration, not a benchmark, so the example shown does not establish measured speed or accuracy.

““Speak it messy. Send it clean.””

— BetterWispr website

Accuracy and Compatibility Still Vary

The supplied product information does not provide independent testing of transcription accuracy, latency or privacy behavior. Performance is expected to vary with the Mac, chosen model, language and recording conditions; the site itself labels its typing illustration as not a benchmark. It is also unclear how well the app performs across different accents, specialist vocabulary and noisy environments.

Automatic insertion may not work in every application or secure field, according to the maker, and supported languages differ across models. The product description also distinguishes local options from external API connections: if users select a third-party provider, dictation may be sent to that provider. The precise handling then depends on the service selected and its terms, which are not detailed in the supplied material.

The launch material does not state a release date, download count, adoption figures or a schedule for future features. It also provides no independent security review or comparison with other dictation products. Those questions remain open beyond the project’s own descriptions.

Testing the App on Your Mac

People interested in trying BetterWispr can consult the project’s website and source repository for installation details and available releases. Before relying on it, users can check the required macOS version, install a compatible model, review the permissions requested and test recognition with their own voice and vocabulary.

Further evidence about the app’s usefulness will depend on hands-on testing and updates from its maintainers and users. Relevant points to watch include model support, language coverage, reliability of text insertion, how local history is handled and whether future releases document changes to data practices. The launch material does not announce a specific next milestone.

Key Questions

Is BetterWispr free?

According to its website, BetterWispr is free and open source under the Apache 2.0 license. It says a subscription, account or API key is not required for its built-in dictation options. Downloaded model files use storage on the Mac.

Which Macs can run BetterWispr?

The stated requirement is macOS 14 or later, and the maker recommends Apple Silicon for local model processing. Available models and performance depend on the computer, operating system and selected language.

Does BetterWispr send speech to the cloud?

The maker says built-in local models transcribe speech on the Mac and that the app has no cloud transcription fallback. Users who choose an external API may send dictation to that provider; that is separate from the described local workflow.

Does it show words while I am speaking?

No. The product information says transcription begins after recording stops. The default workflow is to hold Option + Space while speaking, then release the keys to transcribe.

Which languages and editing features are supported?

Language support depends on the selected model. BetterWispr’s site lists Parakeet support for Japanese and 25 European languages, and broader multilingual recognition through Whisper. It says filler cleanup, app styles and voice commands work in English.

Source: hn