Skip to content
Nikita PrudnikovRU: Читать по-русски
All projects

Platforms: Windows · Linux · macOS

Coming soon

Hyperscribe Companion

Runs Hyperscribe’s speech recognition on your computer’s GPU, for faster captions.

Download

The first public release is being prepared. Until it’s out, you can build it from source.

Build from source
Hyperscribe Companion’s window three times: the welcome screen with the model download, the live transcript, and the compute settings

What it is

An optional desktop accelerator for Hyperscribe. Your phone sends the glasses’ microphone to your computer over Tailscale; Companion recognises the speech on its GPU and sends the captions back to the phone and the lens.

One model on every platform: NVIDIA Nemotron 3.5 ASR, 32 language-locales including Russian, Ukrainian, Chinese, Japanese and Korean, with punctuation. Words reach the phone about a tenth of a second after they are spoken.

Hyperscribe stays usable without it: the phone keeps an offline backup and reconnects when the computer returns. Companion keeps audio and the live transcript in memory only; recordings are saved on the phone.

What it does

  • Any GPU, or none

    NVIDIA through CUDA, AMD through ROCm or Vulkan, Intel through Vulkan, Apple silicon through Metal. With no usable GPU it runs on the CPU, still in real time on a modern machine.

  • Streaming, settled fast

    Words are settled about 0.3 s after they are spoken and are never rewritten after that.

  • Private by default

    It listens only on this computer and on your tailnet through Tailscale’s HTTPS, never on the public internet. Plain local Wi-Fi access is off unless you turn it on.

  • Headless servers too

    On a GPU box without a display, it prints the pairing address and a QR code in the terminal.

Screens

Welcome: This computer does the listening, the speech model and the Download 751 MB and start button
Live: the transcript the phone receives, with the device and the time per refresh
Settings: the model, the compute backend and the server

What you need

  • Windows 10 or 11, Linux (x86_64 or arm64), or a Mac with Apple silicon
  • A GPU is recommended (NVIDIA, AMD, Intel or Apple), but not required
  • About 1 GB of free disk space for the speech model
  • Tailscale on the computer and the phone, and Hyperscribe on your G2

Set it up

Step by step, no coding needed.

  1. Step 1: Set up Hyperscribe first

    Companion speeds up Hyperscribe on your glasses, so install that first and check it works on its own.

    Hyperscribe setup →

  2. Step 2: Install Tailscale

    Tailscale is a free app that links your own devices privately, wherever they are. Install it on the computer and on the phone and sign in with the same account on both.

    Get Tailscale ↗

  3. Step 3: Install Companion

    Download Companion for your system and open it. On Windows the window uses Edge WebView2, which Windows 10 and 11 already have.

    No ready-made download yet: build it from source for now, as described under “For developers”. Build from source →

  4. Step 4: Download the speech model

    Click Download 751 MB and start. Keep the window open while the model downloads, is checked and warms up.

  5. Step 5: Pair your phone

    In Companion, open Pair and wait for an address that starts with wss://. On the phone, open Hyperscribe → Settings → Companion, scan the code, choose When reachable and tap Test connection.

  6. Step 6: Start a recording

    Start on the glasses or the phone. Companion’s Live tab shows the text it sends back. Keep the computer awake and Companion open.

For developers

Build from source

Needs Git, uv and Python 3.11. For NVIDIA, install with '.[cuda]'; for AMD on Linux, run scripts/install-rocm.sh afterwards.

git clone https://github.com/prudnik-web/Hyperscribe-Companion.git
cd Hyperscribe-Companion
uv venv --python 3.11 .venv
source .venv/bin/activate      # Windows: .venv\Scripts\activate
uv pip install -e .            # NVIDIA: uv pip install -e '.[cuda]'
python -m companion

The repository opens with the first public release; these commands work from then on.

What’s new

  1. Coming next

    • One model everywhere: NVIDIA Nemotron 3.5 ASR, with streaming that settles words in about 0.3 s.
    • Windows and Linux support, with CUDA, ROCm, Vulkan, Metal and CPU.
    • A headless mode for machines without a display, and a self-test.

  2. 0.1.0

    • First source release for Apple-silicon Macs, with QR pairing over Tailscale.

More projects

  • Hypernova

    A control room for Claude Code: on your phone, in your menu bar and on your glasses.

  • Hyperscribe

    Live captions on your G2 glasses. Speech recognition stays on your own devices.

  • Hypernavigate

    Offline 3D walking navigation on G2 glasses. Saint Petersburg first.