About these terms
HearPin is developed and operated by TwacQAQ. These terms describe the use of HearPin 1.0 for iPhone and Apple Watch. Questions can be sent to twacQAQ@163.com.
For an App Store download, the software license is governed by Apple’s Standard End User License Agreement, unless a different license is expressly provided in the App Store. These terms supplement that license with product-specific information; they do not replace it or reduce rights that applicable law gives you.
What HearPin provides
HearPin records audio on Apple Watch, transfers clips to a paired iPhone, transcribes them locally, and organizes them into daily Traces with playback and keyword search. It requires compatible devices and the permissions necessary for the features you use.
HearPin 1.0 has no user accounts, in-app purchases, or subscriptions. Any price for downloading the app is shown in the App Store. Apple handles App Store transactions and applicable refund requests under its terms.
Record responsibly
You are responsible for having the right to record, store, and use the voices and other material you capture. Before recording others, give appropriate notice and obtain any consent required by the laws and rules that apply to your situation.
Do not use HearPin for unlawful surveillance, harassment, infringement, or other violations of another person’s privacy or rights. Respect restrictions in workplaces, schools, healthcare settings, and other private or sensitive environments. Recording can be stopped at any time on the watch.
Understand the limitations
Automatic transcripts and their timestamps may be incomplete or inaccurate, particularly with noise, overlapping voices, accents, or languages not well supported by the selected mode. Check the original audio before relying on a transcript for an important decision.
Recording and synchronization can be interrupted by battery levels, storage limits, calls, connectivity, or operating-system behavior. Scheduled tasks may run later than the requested time. HearPin is not an emergency service and does not guarantee continuous recording, complete capture, or an exact transcript.
Your content and storage
You retain any rights you hold in your recordings and transcripts. Using HearPin does not transfer ownership of your content to us. The app processes that content locally to provide its features; we do not claim a right to publish or use your recordings for model training.
You control deletion of your local content and your system backups. Confirm that a transfer is complete before removing the app from your watch. Deleting content may be irreversible, and we do not maintain a server copy for recovery. Read the Privacy Policy for the distinction between local data, pending watch clips, and system backups.
Privacy
Our Privacy Policy explains local data processing, support email, website hosting, and your choices.
Open-source licenses and third-party notices
HearPin includes third-party software and speech models. The following credits and license notices apply to those components. Their licenses govern the relevant components; these terms and the app’s general license do not replace or restrict rights granted by those licenses. The authors and contributors do not endorse HearPin.
Parakeet TDT 0.6B v3 — NVIDIA; CC BY 4.0. HearPin uses the Core ML conversion provided by FluidInference. The conversion changes the model’s deployment format and uses mixed-precision artifacts; HearPin does not further train or fine-tune this model.
Original model and author · Core ML conversion · CC BY 4.0 license
SenseVoiceSmall — FunASR / FunAudioLLM, Alibaba Group; FunASR Model Open Source License Agreement v1.1. Copyright (C) [2023-2028] [Alibaba Group]. All rights reserved. HearPin uses FluidInference’s Core ML conversion with int8 weight quantization. The original model name is retained; HearPin does not further train or fine-tune the model. The model weights are governed by this model-specific agreement.
Original model and author · Core ML conversion · Model license (English) · Original bilingual license
Whisper small — OpenAI; MIT. Copyright (c) 2022 OpenAI. HearPin includes the quantized Core ML conversion distributed by Argmax. The separately bundled tokenizer files from the OpenAI / Hugging Face model repository are distributed under Apache 2.0. HearPin does not further train or fine-tune the model.
Original project · Argmax conversion · Tokenizer source · OpenAI MIT license · Argmax MIT license · Apache 2.0 license
Silero VAD v6.2.1 — Silero Team; MIT. Copyright (c) 2020-present Silero Team. HearPin uses the Core ML conversion distributed by FluidInference for voice activity detection.
Original project · Core ML conversion · MIT license and copyright notice
FluidAudio 0.18.0 — Fluid Inference; Apache 2.0. Provides on-device audio processing and model inference.
WhisperKit / ArgmaxCore 1.1.1 — Argmax; MIT. Copyright (c) 2024 argmax, inc. Provides local Whisper inference and supporting utilities.
Project source · MIT license and copyright notice · Included third-party notices
GRDB.swift 7.11.1 — Gwendal Roué; MIT. Copyright (C) 2015-2025 Gwendal Roué. Provides local database storage.
swift-transformers portions included in ArgmaxCore — Hugging Face; Apache 2.0. Copyright 2022 Hugging Face SAS. These portions were modified by Argmax, Inc. The upstream attribution and modification notice is reproduced in the linked notice file.
Original project · Attribution, modification notice, and Apache 2.0 license
fastcluster portions included in FluidAudio — Daniel Müllner and Google Inc.; BSD 2-Clause. Copyright until package version 1.1.23: © 2011 Daniel Müllner. All changes from version 1.1.24 on: © Google Inc. All rights reserved.
Complete permission notices, warranty disclaimers, and liability terms are provided in the linked license texts and notices. Except where applicable law requires otherwise, these components are supplied on the terms stated by their respective licensors. Questions about these notices can be sent to twacQAQ@163.com.
Updates and your legal rights
We may update the app and these terms as features or requirements change. We will publish updated terms with a revised effective date and provide appropriate notice of material changes. These terms do not promise that every device or operating-system version will remain supported indefinitely.
Nothing here excludes or limits liability, warranties, remedies, or consumer rights where doing so would be unlawful. The warranty and liability provisions of the applicable software license remain subject to those mandatory protections.
Contact
For questions about these terms or using HearPin, contact TwacQAQ at twacQAQ@163.com. For practical help, visit Technical Support.