How to add subtitles to a video
- Drop your video (MP4, MOV, WEBM, MKV) or audio file above.
- Choose speed vs accuracy. Fast is best for quick drafts and older devices, Balanced suits most videos, and Accurate gives the best results for difficult audio. The model downloads once and is cached for next time.
- Wait for the transcript. You’ll see the text appear as each chunk is processed.
- Edit. Fix any mistakes directly in the transcript, adjust timings, split long captions or merge short ones. The video preview shows your captions live as you edit. Find and replace fixes a misheard name everywhere at once.
- Export. Download SRT or VTT to upload alongside your video, or open Burn captions into the video to make an MP4 with styled captions.
Styling burned-in captions
Captions on social video work best when they’re big, bold and high-contrast. Choose a font and size, text and outline colours, and position (bottom, middle or top — middle works well for vertical video where the bottom is covered by app buttons). Turn on word highlighting for the popular karaoke style where each word lights up as it’s spoken. A background box helps readability over busy footage.
Have only audio, such as a podcast clip? The tool can make a captioned MP4 on a plain background, ready for social media.
Why captions matter
- Most social video is watched muted. Captions keep people watching when the sound is off.
- Accessibility. Captions make your content usable for deaf and hard-of-hearing viewers, and are required for many public-sector and education videos.
- Search and reach. Caption files give platforms the text of your video, which helps it get found.
- Comprehension. Viewers watching in a second language, in noisy places or with strong accents follow captioned video more easily.
Good-practice tips
Keep captions to two lines of about 42 characters, and on screen for at least a second. Split captions at natural pauses rather than mid-phrase. Include important sounds in brackets, like [music] or [laughter], for accessibility. Always proofread names, numbers and technical terms — speech recognition is good, but it isn’t a substitute for a quick human check.
Private by design
Online caption services upload your video to their servers, often keep it, and usually add a watermark or a monthly minute limit. Here the AI model is downloaded to your browser and runs locally, so unreleased videos, client work and personal recordings stay on your device. Need a text transcript rather than captions? Use the transcription tool.
Frequently asked questions
How accurate are the subtitles?
For clear speech in a quiet room, the Balanced and Accurate models are usually more than 90% accurate. Accents, background music, crosstalk and technical words reduce accuracy, so always read through and fix the captions in the editor before publishing.
Is my video uploaded?
No. The Whisper speech recognition model runs inside your browser, and captions are burned in with ffmpeg compiled to WebAssembly. Your video never leaves your device.
What languages are supported?
Whisper understands around 99 languages. Leave the language on Auto-detect, or pick it from the list for better results with short clips.
What's the difference between SRT, VTT and burned-in captions?
SRT and VTT are separate caption files that viewers can turn on and off; YouTube, Vimeo, LinkedIn and most video players accept them. Burned-in (open) captions are drawn into the video pixels, which is what you want for Instagram Reels, TikTok and Shorts, where many people watch with the sound off.
Do you add a watermark?
Never. The MP4 you download contains only your video and your captions.
Why is it slow on my computer?
Speech recognition is heavy work. With WebGPU (recent Chrome and Edge), it runs on your graphics chip and is several times faster. Without it, the tool uses your CPU. Choose the Fast model for long files on older devices.
How long can my video be?
There's no fixed limit; audio is processed in 5-minute chunks. Files over about 20 minutes can take a while on slower devices, and burning captions into very large videos (over 500 MB) may exceed the browser's memory.
Can I edit the timing?
Yes. Each caption has editable start and end times, and you can split, merge, add or delete captions. Click the play button next to a caption to hear it.