Model or dataset
umlx5h/LLPlayer avatar
umlx5h/LLPlayer

LLPlayer: dual subtitles, Whisper ASR, and a redistributable you must install first

The media player for language learning, with dual subtitles, AI-generated subtitles, real-time translation, and more!

4,725 stars306 forksC#GPL-3.0

At a glance

What is it?
LLPlayer is a C#/WPF Windows player built around one job, showing a subtitle in the language you are learning and a translation in your own at the same time. Its subtitle pipeline is the product, and almost all of its rough edges live in the dependencies that pipeline needs.
Who is it for?
LLPlayer fits a Windows user studying a language who wants a subtitle in the learning language and a translation in their own on screen at once, and who is willing to install the VC++ redistributable before touching ASR or OCR. Skip it if you are on macOS or Linux, or if a configuration that survives upgrades matters more to you than subtitle features, since the project declares itself Beta and warns that configuration files may not be backward compatible.
Can I use it commercially?
Yes, with conditions. GPL-3.0 is a copyleft licence: if you distribute software that includes it, you must release that software's source code under the same licence. Running it internally without distributing it does not trigger that obligation.
Is it still maintained?
Yes. The repository last received commits 10 days ago.
What is it written in?
Mainly C#, according to GitHub's language statistics.

Answers come from the project's GitHub data, last synced on November 15, 2026, and from our analysis. They are not legal advice.

From daily snapshots

Star history

  • umlx5h/LLPlayer4,725 ★ · +613
Star history
Repository2026-09-102026-11-15
umlx5h/LLPlayer4.1k4.7k

613 Stars in the last 66 days, measured from the daily GitHub snapshot. Compare with other projects

Editorial analysis

The two CC buttons decide which language you are studying

LLPlayer puts two subtitles on screen at once and makes you assign them a job. The bottom seek bar carries two `CC` buttons. The left one is the primary subtitle and the right one is the secondary subtitle. The instructions tell you to set your learning language for the primary and your native language for the secondary.

Both text subtitles and bitmap subtitles are supported in this dual display. The size and position of both subtitles can be adjusted, so the lower track does not have to sit under the upper one by default.

Word lookup runs against subtitle text, and the Subtitles Sidebar supports text and bitmap along with seek and word lookup, plus incremental subtitle search. The practical effect is that a lesson becomes two readable lines at once instead of one line plus a separate dictionary window.

The missing redistributable surfaces only when ASR or OCR runs

Two prerequisites are listed, and they fail in very different ways. The .NET Desktop Runtime 10 is handled politely: if it is not installed, an installer dialog will appear.

The Microsoft Visual C++ Redistributable Version >= 2022 is a different story. It is listed as needed for Whisper ASR and for Tesseract OCR, and the note is plain: if it is not installed the app will launch but will crash when ASR or OCR is enabled. That is a delayed failure rather than a startup check, so a fresh install can look healthy for minutes before the first transcription attempt takes the process down.

For anyone deploying this on a locked down machine, the consequence is that you cannot validate the install by launching it. The redistributable has to be part of provisioning, done ahead of time, because nothing in the startup path reports it missing until you switch on the two features that need it.

Two Whisper engines, two different download flows

Subtitle generation is powered by OpenAI Whisper, and two engines are supported: whisper.cpp and faster-whisper. They differ in how you obtain them, and both flows are spelled out.

For whisper.cpp you go to the `Subtitles > ASR` section and download the models yourself. You choose from a variety of models, and the trade off is stated directly: the larger the size, the higher the load and accuracy. Models with `En` endings are only available in English.

For faster-whisper you download the engine instead of the models from settings. The selected model is then downloaded automatically the first time, so no prior download is required. That removes a setup step but moves the download into the first run, which means your first playback session pays for it.

`Audio Language` lets you set the language of the video manually, and the default is auto detection. Having a CUDA driver will make generating subtitles faster.

Target language is yours to set, source language is not

Translation is configured from the `Subtitles > Translate` section, where the `Target Language` sits at the top. You set this to your native language. The `source language` is detected automatically, so only one side of the pair is manual.

The default translation engine is `GoogleV1`. Switching engines is a separate step: you will need to configure another engine in settings, and the details live in a wiki page that lists the supported ones, including Google, DeepL, Ollama, LM Studio, and OpenAI. Several of those run on your own machine, so the choice also decides whether subtitle text leaves the machine.

Alongside the plain translation path there is context aware translation, which recognizes the context of subtitles using an LLM to raise accuracy. That is a separate behavior from the default engine, so turning it on is a decision about cost and about where text is sent.

Bitmap subtitles become text through OCR, and Microsoft OCR is the other engine

Real time OCR subtitles convert bitmap subtitles to text subtitles in real time, powered by Tesseract OCR and Microsoft OCR. This is the feature that turns an image based subtitle track into something you can search, seek, and look up.

Subtitles seeking for any format means the format of the track you loaded does not decide whether you can jump to a line.

There is also a Built in Subtitles Downloader supporting opensubtitles.org, and integration with browser extensions such as Yomitan and 10ten. Word lookup and browser searches act on subtitle text, with the search site completely customizable and reachable from the context menu of a word.

Online video playback arrives through yt-dlp, not a stream engine

Playing online video is a yt-dlp integration rather than a streaming engine built into the player. Any online video can be played back in real time, with AI subtitle generation and word lookups applied to it.

The input paths are stated plainly. You can play a video from the context menu or by dropping the file. For online videos you can paste the URL with `CTRL+V` or use the context menu. Adding external subtitles is done in the same way as adding videos, either by dragging or from the context menu.

The consequence for readers is that the online feature inherits whatever yt-dlp can and cannot retrieve. A video that yt-dlp cannot resolve does not become playable here. Local files have no such dependency, which makes them the more predictable half of the feature.

Beta means configuration files may not survive an upgrade

The development status is stated as `Beta`, and the project is unusually direct about what that costs. It has not yet been tested by enough users and may be unstable. Significant changes may be made to the UI and settings. Breaking changes will be made actively during version `0.X.X`, and configuration files may not be backward compatible when updated.

That last line is the one to plan around. If you have tuned subtitle placement, shortcuts, or engine settings, an update is allowed to leave those files unreadable by the new build.

The release history is consistent with that posture: v0.2.1 on 2025-04-12, v0.2.2 on 2025-05-24, and v0.3.0 on 2026-04-19. The last push landed on 2026-11-01 and the project is not archived, so work continues even while the version number sits still for months.

Windows x64 is the only platform named, and the build is a solution file

The requirements name two operating systems, Windows 10 x64 Version 1903 later and Windows 11 x64. Nothing else appears in the list, and the build instructions assume the same world: install Visual Studio or JetBrains Rider, then open the solution.

bash
$ git clone [email protected]:umlx5h/LLPlayer.git

Open `LLPlayer.slnx` from Visual Studio or JetBrains Rider using the IDE's Open Project or Solution command. The solution file is not a shell executable.

You then select the `LLPlayer` project and build and run it. The repository layout backs this up: `LLPlayer.slnx`, `Directory.Build.targets`, `LLPlayer/`, `FlyleafLib/`, `FlyleafLibTests/`, `Plugins/`, `FFmpeg/`, and `WpfColorFontDialog/` all sit at the top level.

So on macOS or Linux there is no documented build path, and the reason is structural rather than policy. The player is written in C#/WPF, not C, which the project offers as the reason customization is easy. That same choice is what ties the codebase to Windows.

Editorial conclusion

LLPlayer fits a Windows user studying a language who wants a subtitle in the learning language and a translation in their own on screen at once, and who is willing to install the VC++ redistributable before touching ASR or OCR. Skip it if you are on macOS or Linux, or if a configuration that survives upgrades matters more to you than subtitle features, since the project declares itself Beta and warns that configuration files may not be backward compatible. Before you rely on it, provision the redistributable up front and keep a copy of your settings if you have tuned them.

Frequently asked questions

Can LLPlayer show two subtitles at once?

Yes. Two subtitles can be displayed simultaneously, and both text subtitles and bitmap subtitles are supported. The left `CC` button on the seek bar is the primary subtitle and the right one is the secondary subtitle.

How do I use LLPlayer for subtitles and translation?

Download a build from the releases page, open `LLPlayer.exe`, press `CTRL+.` to open settings, download a Whisper model under `Subtitles > ASR`, set `Target Language` under `Subtitles > Translate`, then play a video from the context menu or by dropping it. Press `F1` for the CheatSheet.

What is an alternative to LLPlayer?

Its stated guiding principle is to be a specialized player for language learning rather than a general purpose player, and the project is released under GPL-3.0 with the source in the repository.

Is LLPlayer suited to learning with or without subtitles?

The dual subtitle setup assumes watching with them. Set your learning language as the primary subtitle and your native language as the secondary one.

Official sources

  1. License: GPL-3.0
  2. Project website
  3. README
  4. Releases
  5. umlx5h/LLPlayer on GitHub
Add this badge to your README

If you maintain this project, the badge below links readers to this analysis and shows its maintenance status from the daily GitHub snapshot. Paste the markdown into your README; add ?metric=license or ?metric=stars to the image URL for a different field.

Add this badge to your README

markdown
[![Hysen Labs](https://hysenlabs.com/badge/umlx5h-llplayer.svg)](https://hysenlabs.com/projects/umlx5h-llplayer)

Star history chart for your README

markdown
[![Star history](https://hysenlabs.com/chart/umlx5h-llplayer.svg)](https://hysenlabs.com/projects/umlx5h-llplayer)

Was this useful?