How do I generate subtitles automatically?
Automatic subtitle generation means turning a video's speech into timed text without typing it yourself. It is the only option when no subtitle file exists, and it is increasingly good enough to watch with. The method differs slightly for web video and local files.
What automatic subtitle generation means
Speech recognition listens to the audio, produces text with timestamps, and those segments become subtitles. A translation step can then turn that text into your language.
For web video (YouTube, X and others)
- Install the WatchTranslate browser extension (desktop Chrome).
- Open the video and pick your subtitle language.
- For audio recognition: Source language defaults to automatic detection; you can also set it manually. Choose the source and languages, then explicitly start translation. Press play — the extension reads the audio, generates subtitles and translates them on the player.
For local and downloaded files
- Open the WatchTranslate local player in your browser.
- Drop in the video and choose your target language.
- For audio recognition: Source language defaults to automatic detection; you can also set it manually. Choose the source and languages, then explicitly start translation. Press play — subtitles are generated and translated as it runs, in about 30-second segments.
Accuracy: what affects it
Recognition quality depends on the audio: clear single-speaker speech works best, while heavy background music, strong accents or several overlapping voices reduce accuracy. This is a limitation of transcription, not of translation.
When a manual subtitle file is still better
If an official subtitle file exists and matches your release, using it gives perfect timing and speaker labels. Automatic generation is for when that file does not exist or does not match.
What it does and doesn't do
- Web video (YouTube, X and others) and local MP4/MKV files
- Generates subtitles from audio, then translates them
- Bilingual original + translated display
- Failed runs do not consume credits
- No live streams
FAQ
Can subtitles be generated with no subtitle file?
Yes — that is exactly what speech recognition does. The audio is transcribed into timed text, which becomes the subtitle track.
Is it accurate enough to watch?
For clear speech it is usually good enough to follow; noisy audio, music or overlapping speakers lower accuracy. The original audio still plays, so you can hear what is being said.
Does it work for local files?
Yes. The local player generates and translates subtitles for files on your computer or phone, without uploading the whole video.
Does it cost anything?
Signup includes one-time free credits. After that it is usage-based, at about 1 credit per minute for subtitles generated from audio.