Twinsubs
VideoCaptioner alternative for Mac: bilingual subtitles made on your own Mac
VideoCaptioner is an open-source (GPL-3.0) subtitle pipeline that runs recognition, line breaking, optimization, translation and burn-in in one command, and its strengths are that it is free and customizable. Twinsubs is a ready-to-run Mac app with nothing to configure. If you like tinkering and want scripted batches, choose VideoCaptioner; if you want to drop a file in and get subtitles, plus live translation, choose Twinsubs.
Where they differ
| Item | Twinsubs | VideoCaptioner |
|---|---|---|
| Form | Mac app (on the Mac App Store) | Open-source desktop tool and CLI (Windows installer, macOS one-line script) |
| Where it runs | On your Mac, nothing uploaded | Whether the free recognition and translation path goes online: not verified (the README does not say); the large-model features use your own API key |
| Price | Free for the first 10 minutes of each video and 10 minutes of live translation per day; one-time purchase removes the limits | Free and open source (GPL-3.0); model features cost whatever your provider charges |
| Bilingual subtitles | Yes, translation above the original (order can be swapped) | Yes |
| Live translation | Yes, the sound playing on your Mac; works offline once the language packs are downloaded | No (not on the page) |
| Export SRT | Yes, a bilingual file and a target-only file | Yes (outputs SRT) |
“Not verified” means the official page has no clear statement, or the answer sits in dynamic content we could not read; it does not mean the feature is missing. We did not sign up for any account or try any paid feature.
When to choose VideoCaptioner
- You want free software with public code. The GitHub page showed about 16.2k stars the day we read it, under GPL-3.0, so you can change it.
- You want a large model doing line breaking and translation polish. The README says any OpenAI-compatible provider works and switching is a configuration change.
- You want scripts. The command line covers transcription, translation, burn-in and downloading online video, which suits batches and automation.
When to choose Twinsubs
- You do not want to set up an environment. Twinsubs is a Mac app: download the speech model once and drop files in, with no Python and no keys.
- You want live translation of the sound on your Mac, which the VideoCaptioner page does not mention.
- You want recognition and translation to stay on your own Mac. Twinsubs does both on-device; for the free path in VideoCaptioner the README does not say.
Limits of Twinsubs, stated up front
- System requirement: macOS 26 or later, Apple silicon.
- It outputs SRT only: no subtitle burn-in, no built-in editor, no speaker separation, no ASS, and it cannot import an existing SRT and translate it.
- Machine recognition and translation make mistakes; with heavy accents, loud music or many specialist terms, check the result against the video.
- Which language pairs work depends on Apple's on-device translation; there are 12 source languages and 20 target languages.
Moving from VideoCaptioner
VideoCaptioner outputs SRT, so videos you have already processed do not need redoing. To compare, drop the same video into Twinsubs and look at recognition and line breaking: Twinsubs cuts by clause, Chinese lines stay within 20 characters, and English within two lines of 42. If you rely on command-line scripts, keep VideoCaptioner for batches and use Twinsubs for day-to-day single videos and live translation.
A few notes
The VideoCaptioner README says the free features (Bijian speech recognition and Bing or Google translation) need no configuration, but it does not say whether those paths send audio or text to the provider, and we will not decide that for it, so it is marked not verified. The star count changes over time.
How to check this yourself
- Read the other product's pricing pagePrices, plans and free allowances change; trust the page on the day you buy.
- Run the same clip through bothPick a clip with an accent and some technical terms; it tells you more than a spec sheet. The free version of Twinsubs covers the first 10 minutes of each video.
- Check the export formatConfirm you can export SRT and that your usual editor imports it without errors.
- Check where the data goesRead the other product's privacy policy to see whether video or audio is uploaded. Ours is the privacy policy.
Sources (verified 2026-10-16)
- VideoCaptioner on GitHub · 2026-10-16
“Not verified” means the official page has no clear statement, or the answer sits in dynamic content we could not read; it does not mean the feature is missing. We did not sign up for any account or try any paid feature.
FAQ
Is VideoCaptioner free?
The software is free and open source under GPL-3.0. The README says the free features need no configuration; the large-model features need your own API key, and the cost is set by the provider you choose.
Does VideoCaptioner run on a Mac?
The README mentions a macOS one-line install script, and it can also be installed with pip as a command line plus desktop app. See the project documentation for system requirements.
How reliable is the information on this page?
Every fact carries an official source and a verification date (2026-10-16), and anything we could not find is marked not verified. Prices and features change, so check the other product's own site before you buy; if you spot a mistake, write to twinsub@gatheon.com and we will correct it after checking.
Updated 2026-10-23. Back to the Twinsubs home page