Terms of Use
1. Acceptance
By accessing any Nihon Kanun K.K. product, you agree to these Terms and the applicable App Store/Apple Media Services terms. If you are under the age required in your region, you must obtain guardian consent.
2. License
We grant a personal, non-transferable, non-exclusive license to use our apps. Reverse engineering, redistribution, sublicensing, or automated scraping of our services is prohibited.
3. Purchases & Subscriptions
All paid features are billed through Apple. Refund requests must follow Apple's policy. Transaction identifiers may be checked to validate active entitlements.
4. Data Handling
Features that rely on cloud processing (e.g., Calendar It! share extension) send user-provided content to trusted LLM providers such as OpenAI or Google Gemini only to generate calendar drafts, then discard it immediately. See our Privacy Policy for details.
5. User Obligations
- Do not interfere with or overload our APIs.
- Do not resell or repackage the service to third parties.
- Respect intellectual property and comply with local laws.
6. Disclaimer & Limitation
The apps are provided "as is." To the maximum extent permitted by law, we disclaim warranties of fitness and are not liable for indirect or consequential damages.
7. Governing Law
These Terms are governed by the laws of Japan. Disputes will be submitted to the Osaka District Court.
8. Updates
We may revise these Terms when services or regulations change. Material changes will be announced on this page.
WatchMemory: Open-source software and models
WatchMemory uses the following components for on-device speaker separation. Their respective licenses apply to those components; the restrictions above do not override rights granted by those licenses.
FluidAudio 0.15.1
FluidInference publishes FluidAudio under the Apache License 2.0. Full license.
Speaker diarization models
The Core ML models converted and optimized by FluidInference are based on pyannote speaker-diarization-community-1 and distributed under CC BY 4.0. Full license ยท Model card and attribution.
Speaker embeddings use work from WeSpeaker; speaker clustering uses work from VBx. The model card contains the component details and research credits. License and model-card files are also included in the app bundle.
For audio, transcription, summarization modes and daily reminders, see the WatchMemory Privacy Policy.