Generate, Edit, and Hardcode Video Subtitles Without Uploading Files
Adding accurate captions to a video usually forces you to choose between two frustrating options: spending hours typing out every spoken sentence by hand, or uploading gigabytes of private footage to a cloud service that slaps a watermark on your export or locks SRT downloads behind a paid subscription.
Our browser-based auto-subtitling studio eliminates both problems. By combining OpenAI's Whisper speech recognition models (running locally via WebAssembly) with hardware-accelerated WebCodecs video rendering, this tool listens to your video's audio track, transcribes or translates the spoken dialogue, lets you fine-tune every timestamp and visual style, and burns the final captions directly into your video—all inside your browser tab with zero server uploads.
How Local Whisper AI Speech Recognition Works in Your Browser
Once you click Choose video or drop an MP4, MOV, WebM, or MKV file into the upload area, the editor inspects your media streams and prepares the Generate subtitles automatically (AI) panel in the sidebar. Instead of sending your audio over the internet, the tool decodes the audio track into 16 kHz mono samples right in your device's memory and slices long recordings at natural quiet pauses (~60-second segments) so words are never cut in half.
Before clicking Generate subtitles, you can tailor the AI engine to match your exact footage:
- Spoken language: Leave it on Auto-detect if your clip contains mixed speech, or manually lock in one of the 16 supported languages—including English, Arabic (العربية), French (Français), Spanish, German, Turkish, Hindi, Chinese, Japanese, and Korean—for faster and cleaner recognition.
- Model quality: Choose between three quantized Whisper models that download once and stay cached in your browser for future visits:
- Fast (≈40 MB): Powered by whisper-tiny. Ideal for clear English voiceovers and quick drafts on lower-end laptops.
- Balanced (≈75 MB) — recommended for Arabic: Powered by whisper-base. Delivers a strong jump in accuracy for non-English languages, accents, and regional dialects without slowing down your computer.
- Most accurate (≈250 MB) — slower: Powered by whisper-small. Best for fast speakers, technical terminology, or background noise where maximum transcription precision matters.
- Task: Select Write the speech as spoken to transcribe captions in the speaker's original language, or switch to Translate the speech to English to automatically convert foreign-language dialogue into English subtitles in a single step.
- Line length: Control how many characters appear per caption block before the tool automatically splits long sentences. Choose Short (32 chars) for vertical reels and TikTok clips, Normal (45 chars) for standard YouTube videos, or Long (72 chars) for widescreen lectures.
Tip for long recordings: To protect your browser's RAM, AI transcription supports videos up to 40 minutes long. If you are working with a full-length lecture or podcast, split it into chapters first using our Video Splitter or trim out unwanted dead air with the Video Trimmer before generating captions.
Full Timeline Control: Import, Edit, Sync, and Export Transcripts
AI gets you 95% of the way there in seconds, but human polish makes captions look professional. Under the Subtitles section, every generated or imported caption appears as an interactive card linked to the live video player:
- Instant Jump & Live Highlighting: Click anywhere on a subtitle card to pause the player and jump directly to that exact millisecond on the canvas. As the video plays, the active line automatically highlights in blue.
- Manual Cues & Existing Files: Already have a caption file? Click Load SRT / VTT file (with automatic UTF-8 and Windows-1256 Arabic encoding fallback) to import it, or click Add a line at current position to insert a new 2.5-second caption block right where the playhead sits.
- Global Timing Offset: If an imported subtitle file is slightly out of sync with your video cut, you don't need to edit 200 timestamps one by one. Enter a positive or negative value in Timing offset (seconds) (for example, +0.5 to delay captions by half a second, or -1.2 to show them earlier) and watch the preview adjust immediately.
- One-Click Text & Subtitle Exports: Even if you don't need hardcoded video captions, you can use this tool purely as a free transcription studio. Click Export SRT for YouTube/Facebook closed captions, Export VTT for HTML5 web players, or Export text to download a clean, timestamp-free plain text (.txt) transcript for blog posts and show notes.
Customizing Subtitle Typography and Burn-In Visual Styles
When you burn subtitles into a video (often called "open captions" or "hardcoded subtitles"), the text becomes a permanent part of the video frames so viewers can read it even on mute or on platforms that don't support external SRT files. Our live HTML5 canvas overlay previews your exact styling in real time, including automatic Right-to-Left (RTL) text direction for Arabic, Persian, Urdu, and Hebrew scripts:
- Font & Weight: Pick from clean sans-serif faces (Arial / Tahoma or dedicated Tahoma for crisp Arabic rendering), classic editorial serif (Georgia (serif)), or technical typewriter style (Courier (monospace)), and toggle the Bold checkbox for extra punch.
- Relative Scaling & Placement: Adjust the Font size slider from 2% to 12% of the video's height so your text scales proportionally whether your video is 720p or 4K. Set the vertical Position to Top, Middle, or Bottom, and fine-tune the Distance from edge (0% to 30%) so captions never collide with lower-third graphics or social media UI buttons.
- Contrast & Background Styles: Make white or colored text readable against bright footage by choosing Outline (draws a crisp stroke around each letter), Box (places a 72% semi-transparent rounded background bar behind each line), Shadow (adds a soft drop shadow), or None, with independent color pickers for both Text color and Outline/box color.
Hardware-Accelerated Video Output & Compression Settings
Under the Output section, you control how the final subtitled video is encoded before clicking Burn subtitles into video:
- Output Format: Choose MP4 (H.264) — most compatible for universal playback across iPhones, Android devices, Premiere Pro, and social platforms, or select WebM (VP9) for superior web compression at roughly 25% smaller bitrates.
- Max Size (Resolution Scaling): Keep the resolution at Original to preserve every pixel, or downscale heavy 4K footage on the fly to a maximum long-edge of 1920 px (1080p), 1280 px (720p), or 854 px (480p).
- Bitrate Quality & Live Size Estimator: Switch between Low, Medium, and High quality presets while watching the Estimated size readout update before you even start encoding. If your final video still needs to fit strict email or Discord upload limits afterward, you can pass it through our dedicated Video Compressor.
Creative Workflows: Pair Subtitled Clips with Other Browser Tools
Once your captions are burned in and you review the side-by-side Original and Result players on the completion screen, you can take your content even further without leaving your browser:
- Adjust Pacing Before Subtitling: If a tutorial or presentation feels too slow, speed it up first using our Video Speed Changer so the AI generates timestamps that match your final fast-paced cut.
- Create Silent Social Teasers: Turn a 5-second subtitled punchline or reaction clip into a lightweight looping animation for emails and documentation with the Video to GIF Converter, then optimize its file weight in our Animated GIF Compressor.
- Design Custom Video Thumbnails: Grab a clean frame from your video and prepare eye-catching cover dimensions using our Image Resizer before publishing your captioned video to YouTube or Reels.