An honest comparison with existing voice control and dictation tools.
| Samsara | Dragon Professional | Talon Voice | Windows Voice Access | OpenWhispr | |
|---|---|---|---|---|---|
| Price | Free, AGPL-3.0; optional third-party services may charge. | Not publicly verified | Not publicly verified | Not publicly verified | Not publicly verified |
| Platform | Windows 10/11 | Not publicly verified | Not publicly verified | Not publicly verified | Not publicly verified |
| Works offline | ✓ Core dictation and wake processing are local; downloads and optional features can use network. | Not publicly verified | Not publicly verified | Not publicly verified | Not publicly verified |
| Speech engine | Whisper via faster-whisper; optional NVIDIA CUDA acceleration. | Not publicly verified | Not publicly verified | Not publicly verified | Not publicly verified |
| Voice commands | Extensive: built-ins and plugins | Not publicly verified | Not publicly verified | Not publicly verified | Not publicly verified |
| Custom commands | Python command plugins. | Not publicly verified | Not publicly verified | Not publicly verified | Not publicly verified |
| Dictation | Hold-to-dictate, Continuous, and Hands Free dictation. | Not publicly verified | Not publicly verified | Not publicly verified | Not publicly verified |
| Training required | None | Not publicly verified | Not publicly verified | Not publicly verified | Not publicly verified |
| Wake word | ✓ Configurable (OpenWakeWord or transcript fallback) | Not publicly verified | Not publicly verified | Not publicly verified | Not publicly verified |
| AI assistant | ✓ Ava: local Ollama or user-configured cloud LLM. | Not publicly verified | Not publicly verified | Not publicly verified | Not publicly verified |
| Text selection | Voice markers and selection commands. | Not publicly verified | Not publicly verified | Not publicly verified | Not publicly verified |
| GPU acceleration | ✓ Optional NVIDIA CUDA acceleration. | Not publicly verified | Not publicly verified | Not publicly verified | Not publicly verified |
| Privacy | No telemetry; optional network paths disclosed | Not publicly verified | Not publicly verified | Not publicly verified | Not publicly verified |
| Open source | ✓ AGPL-3.0. | Not publicly verified | Not publicly verified | Not publicly verified | Not publicly verified |
| Best for | Accessibility + desktop control | Not publicly verified | Not publicly verified | Not publicly verified | Not publicly verified |
Dragon Professional · Talon Voice · Windows Voice Access · OpenWhispr
Samsara uses OpenAI's Whisper speech recognition model, running locally through
faster-whisper — a CTranslate2-optimized build. CTranslate2 is a C++
inference engine. The standard release runs on CPU; the optional, separately downloaded CUDA runtime
can substantially reduce transcription latency on a compatible NVIDIA GPU. Actual speed depends on the
model, processor or GPU, audio length, and settings.
Wake-word profiles can use OpenWakeWord ONNX models. Profiles without a compatible model
can use the Whisper transcript fallback, so resource use depends on the configured profile and mode.
Voice activity detection uses Silero VAD, a lightweight neural network that classifies
audio as speech or silence before Whisper runs. This prevents wasted cycles on background noise.
The AI assistant (Ava) runs on local Ollama models or optionally routes to cloud providers
(DeepSeek, OpenAI, Anthropic, OpenRouter) when the user explicitly enables cloud mode and supplies a key.