Features

Here's what EchoVCast does and what to expect.

Local Speech Recognition

Speech recognition runs locally on your machine. An NVIDIA GPU gives the best speed, but CPU mode is available with higher latency. Accuracy varies — clear audio with one speaker works best. Background noise, heavy accents, or fast overlapping speech will cause more errors. Your audio never leaves your computer.

  • GPU acceleration or CPU fallback
  • Dedicated engine for each language
  • Voice activity detection to filter silence
Local
Speech recognition stays on your machine
English Japanese
English Traditional Chinese
Japanese Traditional Chinese

Machine Translation

Transcribed text is sent to our server for machine translation. It supports bidirectional translation between English, Japanese, and Traditional Chinese. Translations are generally good for everyday speech, but can struggle with idioms, cultural references, and ambiguous context.

  • Uses recent context for better coherence
  • Bidirectional language pairs
  • Works best with clear, conversational speech

OBS Integration

EchoVCast writes to OBS-compatible text files and can send closed captions via WebSocket. Some setup is needed on the OBS side — see the instructions page for a walkthrough.

  • OBS text source file output
  • WebSocket closed captions (CEA-608)
  • Separate mic and system audio channels
OBS
Text sources + closed captions
Microphone Input Your voice
System Audio YouTube, Twitch, Discord, etc.

Dual Audio Capture

Capture and process both your microphone and system audio at the same time. Useful for translating livestreams, videos, or voice calls while also capturing your own voice. Both sources run independently with separate language settings.

  • Microphone and system audio at the same time
  • WASAPI loopback for system audio capture
  • Independent language settings per source

And More

Session Transcripts

Export full session logs as TXT, SRT, or VTT subtitle files.

Global Hotkeys

Control translation with keyboard shortcuts while in other apps.

Profanity Filter

Optional word-based filter to keep your stream clean.

Built-in Updater

Check for updates and install new versions directly from the app.

Windows Native

Windows 10/11 only. NVIDIA GPU recommended, with CPU fallback available.

Ready to get started?

EchoVCast is in beta and actively being improved. We'd love your feedback.

View Pricing