Live Speech Recognition

Recognize system audio and display the original text or translation as subtitles

EasyChat can recognize audio playing on your Windows system, translate it, and display the result as subtitles. Ten recognition languages are supported.

How to use it

ASR model required

Currently, the ASR models for real-time speech recognition can be downloaded directly within the app (download source: GitHub). Users in mainland China may experience download failures due to network issues. You can configure a proxy in Settings -> General -> Network Proxy before trying again, or use tools like Watt Toolkit to accelerate GitHub downloads;
Alternatively, you can manually import models downloaded from GitHub, Google Drive, or the group files in our QQ Group.

Under Setting > Speech > ASR models, import the required model folder or supported model archive.
Open Speech Recognition, choose the audio source and recognition options, and start recognition. Enable the Subtitle Window to show the original text or translation over other applications.

The audio source can be limited to a specific process so sounds from other applications are ignored.

Supported languages

10 recognition languages are supported, each with its own ASR model. Please download the corresponding model before using the feature.

Language	Code
English	en-US
Chinese (Simplified)	zh-CN
Japanese	ja-JP
Korean	ko-KR
German	de-DE
Danish	da-DK
Spanish	es-ES
French	fr-FR
Italian	it-IT
Portuguese	pt-PT

Use cases

Videos and live streams without subtitles

Follow foreign-language videos and streams with translated subtitles generated from system audio.

International meetings and online classes

Select the meeting application's audio process to receive focused subtitles without interference from other software.

Podcasts and audio-only content

Turn podcasts, audiobooks, and interviews into readable translated text.

Listening practice

Compare spoken language with real-time source or translated subtitles while studying.