Subtitle Generator
Use the Subtitle Generator to produce subtitles with AI speaker detection, filler removal, and privacy protection. Visual timeline editing keeps every Subtitle Generator export aligned for SRT/VTT.
Why teams choose Harku
Built for professional subtitle workflows
Upload audio or video into the Subtitle Generator, clean filler words, adjust timing on the waveform, and export production-ready SRT/VTT without leaving the editor.
Real speaker names, not Speaker 1/2
Detects self-introductions (“Hi, I'm…”) in English & Chinese to suggest real names.
One-click filler cleanup
Remove 35+ filler words (um, uh, like, 嗯, 这个) while keeping natural pacing.
Off-record privacy switch
Mark sensitive lines once—they stay out of every export by default.
Drag & drop your audio or video file into the Subtitle Generator to start producing subtitles.
The Subtitle Generator supports MP4, MOV, MKV, WebM, MP3, WAV, M4A, AAC, FLAC for instant uploads.
Guests get a 5-minute Subtitle Generator preview. Sign in anytime to unlock full subtitles and exports.
High accuracy
Sign in to enable high accuracy (Pro plan).
Speaker diarization
Sign in to enable speaker diarization (Basic plan or above).
Preview sample subtitles
Subtitle Generator deliverables included with every transcription
Everything the Subtitle Generator produces so you can publish accurate subtitles immediately.
Accurate subtitle segments
Word-level timestamps with speaker labels and paragraph grouping for long-form content.
Waveform editing workspace
Adjust timing visually, split or merge segments, and preview playback in real time.
AI productivity tools
One-click filler removal, smart speaker suggestions, and automatic fact-check highlights.
Multi-format exports
Download SRT, VTT, TXT, Markdown, or open the editor to refine later.
AI Summary
- • Two-minute clip from a product demo where a host and engineer explain how the Harku Subtitle Generator keeps subtitles synced.
- • Focuses on quick upload, waveform review, and exporting SRT/VTT in one Subtitle Generator workflow.
Shown in the editor
- • Speaker rename turns "Speaker 1" into "Maya (Host)" in a single Subtitle Generator click.
- • Filler word cleanup removes "um"/"like" without shifting timestamps so the Subtitle Generator keeps accuracy.
- • Off-record toggle hides an internal pricing note from every Subtitle Generator export.
What you get
- • Polished subtitle segments with timestamps and speaker labels produced by the Subtitle Generator.
- • Waveform-aligned transcript ready for fine-tuning inside the Subtitle Generator.
- • Instant downloads in SRT, VTT, TXT, Markdown, and JSON directly from the Subtitle Generator.
Smart Subtitle Generation
Auto Subtitle Generator
The Subtitle Generator lets teams upload, edit, and export subtitles without juggling multiple tools.
- ⚡Instant Generation
- Upload MP4/MP3, get subtitles in 30 seconds
- 🌍Multi-language
- 50+ languages with auto-detection
- 🆓Free to Start
- 30 min/day, no signup required
- 🔒Privacy First
- Auto-delete in 24h, your data stays private
Smart Features You Won't Find Elsewhere
Professional features designed for serious work
AI speaker detection, filler cleanup, and off-record controls are built directly into the Subtitle Generator workflow.
AI Speaker Detection
Auto-suggests real names from self-introductions
- "Hi, I'm John Smith" → Suggests "John Smith"
- "大家好，我是张伟" → Suggests "张伟"
- Works for Chinese & English names
- One-click to apply suggestions
Unlike other tools that only provide generic 'Speaker 1/2/3' labels, our AI automatically recognizes and suggests real names from introductions, saving hours of manual renaming work.
Filler Word Removal (EN+CN)
One-click cleanup of 35+ filler words for professional transcripts
- English (17): um, uh, like, you know...
- Chinese (18): 嗯, 啊, 这个, 那个...
- Before: Um... like... I think... you know... it's good
- After: I think it's good
Most competitors like Descript only support English filler words. We're the only solution offering comprehensive filler removal for both English and Chinese speakers.
Off-Record Protection
Mark sensitive content that won't appear in exports
- Click segment → Mark as off the record
- Auto-filtered on export
- Perfect for investigative journalism
- Protect sources and sensitive info
An exclusive feature not found in any other subtitle tool. Perfect for journalists, lawyers, and professionals who handle sensitive information.
Frequently Asked Questions
How to create SRT from MP4?＋
Upload your MP4 file, click "Upload to Generate", wait 30 seconds for auto-generation, then click "Export" and select "SRT" format. Your .srt file will download immediately with proper timing and UTF-8 encoding.
What's the difference between SRT and VTT?＋
SRT (SubRip) is the most universal format, compatible with all video players and platforms. VTT (WebVTT) is web-optimized with support for styling and positioning, ideal for HTML5 video players. Both formats are free to export from our tool. When in doubt, choose SRT for maximum compatibility.
How to shift subtitle timing?＋
If your subtitles are out of sync, use the "Time Shift" tool in the editor. Enter an offset in milliseconds (+500ms to delay, -500ms to advance) and click "Apply to All". You can also drag individual segments on the visual timeline for precise adjustments.
Is it really free?＋
Yes! Free users get 30 minutes of transcription per month with all features included: auto-generation, visual editor, AI speaker detection, filler removal, and SRT/VTT export. No watermarks, no signup required. Upgrade anytime for unlimited usage.
Which languages are supported?＋
We support 50+ languages including English, Chinese (Simplified/Traditional), Spanish, French, German, Portuguese, Italian, Japanese, Korean, Russian, Arabic, Hindi, and more. Our AI automatically detects the language.
Can I add subtitles to my video?＋
Yes! After generating subtitles, export the SRT or VTT file and import it into your video editor (Premiere Pro, Final Cut, DaVinci Resolve, etc.) or directly upload to YouTube. For burned-in subtitles, use your video editor's "Burn Subtitles" feature.
How does AI speaker detection work?＋
Our AI listens for self-introduction patterns like "Hi, I'm John" or "大家好，我是张伟" and automatically suggests the speaker's real name. You can review suggestions and apply them with one click. This saves hours of manually renaming "Speaker 1, 2, 3".
What are filler words?＋
Filler words are verbal hesitations like "um", "uh", "like", "嗯", "这个", "那个" that appear in natural speech. Our tool can remove 35+ filler words in English and Chinese with one click, making your transcripts more professional and readable.
What does "off the record" mean?＋
Content marked as "off the record" will be automatically excluded from all exports. This is useful for journalists protecting sources, legal professionals handling confidential information, or anyone who needs to remove sensitive parts of a conversation before publishing.
How long does it take to generate subtitles?＋
Typically 30-60 seconds for a 10-minute file. Processing time depends on file length and current server load. You'll see real-time progress updates.
What file formats can I upload?＋
Videos: MP4, MOV, AVI, MKV, WebM Audio: MP3, WAV, M4A, AAC, FLAC Maximum file size: 500MB Maximum duration: 30 minutes (free), 4 hours (paid)
Ready to create professional subtitles?
Start free — no signup required, no watermarks.