Is it free and open-source?

Yes. LiveAudio is free and open-source under the MIT license. No subscription and no API key.

Low, tunable latency — well under a second on a typical setup. Profiles trade latency against accuracy and GPU load.

PLYNTE LIVEAUDIO v1.5.0 · RUST + TAURI 2 · WINDOWS X64

Real-time local captions for your OBS.No cloud, no subscription, no API key.

LiveAudio is a free, open-source (MIT) app that generates real-time Whisper speech captions 100% locally and streams them to OBS over a local WebSocket. v1.5.0 is rewritten in Rust with a Tauri 2 desktop shell: faster startup, lower RAM, memory-safe transcription — and no Python runtime to download.

Download LiveAudio v1.5.0 (free)How it works

Rust + Tauri 2
100% local
MIT open-source

LiveAudio product mark with the LiveAudio name and Plynte attribution.

Rust + Tauri 2
100% local
MIT open-source
No cloud
No API key

LIVE DEMO

See LiveAudio running live.

The current v1.5.0 interface in one screenshot — plus an archived recording of the deprecated Python version below.

real capture

LiveAudio v1.5.0 desktop showing the Advanced tab with Whisper turbo on CUDA, voice-activity tuning, live subtitle feed and OBS Studio integration. — v1.5.0 (Rust + Tauri 2) — the current interface.

archived capture · deprecated Python version

Archived recording of the deprecated Python version — the v1.5.0 screenshot above is current. A new Tauri video is coming soon.

// why LiveAudio

Straight answers to the real objections.

// 100% local

Is it really 100% local?

All processing happens on your machine, with no telemetry. Internet is needed only on first run to download transcription models; after that it runs fully offline. v1.5.0 ships as a single native binary — no Python runtime to provision.

rust native binary · no telemetry · offline after first run

// linux

Does it work on Linux?

v1.5.0 focuses on Windows 10/11 x64. Linux users stay on the legacy Python v1.2.7 tarball (microphone capture, needs libportaudio2) until a native Linux build ships. System-audio loopback (WASAPI) is Windows-only.

v1.5.0 windows x64 · linux legacy v1.2.7

loopback Windows-only

// gpu

Do I need a GPU?

No. The CPU works. NVIDIA CUDA is optional but recommended: it is auto-detected and needs driver ≥ 525 and VRAM ≥ 4 GiB.

cpu works · cuda driver ≥ 525 · vram ≥ 4 GiB

// the signal path

From microphone to OBS, locally.

Capture audio, keep real speech, decode it locally, then send captions to OBS. Read the docs when you need the implementation details.

Read the signal-path docs

Capture
Choose a microphone or supported system audio input.
mic · system loopback (windows)
Gate
Silero VAD keeps silence out before Whisper decodes.
silero-vad · onset pre-roll + threshold
Decode
Whisper turns speech into text on CPU or optional CUDA.
whisper · tiny / base / small / turbo
Broadcast
Local subtitle JSON reaches OBS or another localhost client.
ws://127.0.0.1:8765

// choose a path

See the product, install it, or wire it into OBS.

The home page gives you the decision. These routes carry the details without making you scroll through every possible configuration.

Product Tour

Walk through real released-app states, from first run to diagnostics.

Open the tour

Getting Started

Install, choose a profile, and start the local caption engine.

Start setup

OBS Setup & Docs

Connect the local subtitle output to OBS, then keep the technical reference nearby.

Set up OBS Browse docs

// download

Local captions in your OBS in minutes.

Free & open-source (MIT). No subscription, no API key. You only pay your own electricity — hardware not included.