akous

Your Private
Soundscape.

Generate infinite, immersive ambient environments from text prompts using on-device AI. No cloud required. Powered by Apple Silicon.

Akous Home Screen

100% On-Device AI

Powered by an ARC-distilled Diffusion Transformer running entirely on your iPhone's Neural Engine. Describe any scene — "ocean waves on a rocky shore" — and akous creates it locally in seconds. No cloud, zero latency.

Generate Tab

Private by Design

Your prompts and audio never leave your device. No analytics, no tracking, no data collection. Just you and your imagination.

Pro Built-in Mixer

Shape your soundscapes with a 5-band EQ, dual LFOs, real-time granular synthesis, binaural beats, and a noise generator. Infinite sonic exploration.

9-Stage Pipeline

Every output passes through a professional post-processing chain: DC removal, glitch repair, edge fading, HPF/LPF, peak limiter, de-esser, and safety soft-clip.

Technical Specs

Model Architecture ARC-Distilled Diffusion Transformer (DiT), VAE Decoder, T5 Text Encoder
Generation Speed 15-30 seconds on Apple Silicon (A12 Bionic and newer)
Audio Output High-fidelity 11s clips with seamless edge-faded looping; export lengths from 1m up to 8h (WAV)
Hardware Acceleration Fully optimized for Apple Neural Engine (ANE) and Metal GPU
Customization Adaptive sampling, customizable steps (4-32), noise temperature (0.6-1.0), and seed control

Privacy & Security

akous is built on the principle that AI should enhance your capabilities without compromising your privacy. Here's how we deliver on that promise:

No Data Collection

  • No Personal Information: We do not collect, store, or transmit any personally identifiable information (PII).
  • No Analytics: Zero third-party analytics, telemetry, or tracking SDKs are integrated.
  • No Cloud Storage: All processing occurs locally on your device — your prompts and generated audio never leave your iPhone or iPad.
  • No Internet Permissions: The app does not require network access for core functionality.

Data Ownership & Control

  • Your Data, Your Control: All saved soundscapes and settings are stored locally in the app's iOS sandbox.
  • No Background Transmission: The app cannot send data without your explicit action.
  • Local Model Storage: The CoreML models are bundled directly with the app.

Technical Privacy Measures

  • Sandboxed Execution: Models run in iOS's secure sandbox with no access to other apps or system data.
  • In-Memory Processing: All inference calculations happen in RAM and are cleared when the app closes.

Transparency

akous is built on the foundation of CoreML and Apple's Privacy-first philosophy. The app's privacy policy is simple: we do not touch your data. If you are concerned about privacy, akous gives you the most secure environment possible.

Designed for Your Focus

The Architect

Deep, non-repetitive focus-scapes for programmers and system designers who need to stay in "The Flow" for hours.

The Dreamer

Ethereal, evolving textures designed to lower cortisol and facilitate deep, restorative sleep and lucid dreaming.

The Zen Master

Custom binaural beats and frequency-aligned soundscapes for meditation, breathwork, and mindfulness practitioners.

The Nomad

Offline generation for travelers and commuters who need a private acoustic sanctuary in loud, public environments.

Frequently Asked Questions

akous utilizes Apple's CoreML and Metal frameworks to run a 341-million parameter Diffusion Transformer directly on your iPhone's Neural Engine. By distilling the model for mobile, we've achieved professional-grade synthesis that doesn't require a server farm or internet connection.

No. Once the app is installed, all soundscape generation works entirely offline. The CoreML models are bundled directly within the app. An internet connection is only required for the initial download and App Store transactions.

ARC (Adversarial Robustness & Convergence) distillation is a state-of-the-art training technique that allows the AI to produce high-fidelity results in significantly fewer steps. This is what enables akous to generate complex soundscapes in seconds rather than minutes.

akous requires iOS 17.0+ and a device with Apple A12 Bionic (iPhone XS/iPad Air 3rd gen) or later. For the best experience and fastest generation times, we recommend an iPhone 13 or newer.

Absolutely not. Because all processing happens on your device, we never see your text prompts or your generated audio. Your creativity is yours alone, and we have no way to access it for training or any other purpose.

Yes! You can export any generated soundscape as a high-quality WAV file via the iOS system share sheet. You can save it to the Files app, send it via Messages, or import it into pro audio apps. Exports can range from 1 minute up to 8 hours.

Joshua Hrisko

Maker Portal is a blog-centric company intended for young innovators interested in real-world applications to engineering. Resources include: physical products, mobile applications, software development, e-learning, and blog-style article writing. The maker-based approach is explored using written articles with topics ranging from Raspberry Pi, heat transfer, acoustics, robotics, data analysis, Arduino, sensor design, Python programming, and much more. Difficulty levels range depending on the topic and there is extensive focus on open-source software implementation, however, there will be articles with a focus on software design as well. The intention is to demonstrate applications of engineering that are repeatable at the intermediate level without requiring colossal resources. 

https://makersportal.com/
Next
Next

itria