Burn word-by-word animated captions onto any video — the TikTok-viral style that increases watch time. Local Whisper transcription, three preset styles, no watermark.
Most captioning tools display a full sentence at once. ViralMint highlights one or two words at a time, synced to the speaker — the same style that drives 80%+ of viral TikToks and YouTube Shorts. Three presets (viral, classic, bold) cover the full range without manual styling.
Transcription runs on your machine via faster-whisper (CPU int8). Your video file never leaves your computer. Word-level timestamps come back in seconds for short clips and a couple of minutes for a 30-minute podcast — and crucially, accurate enough to drive caption timing without manual cleanup.
Submagic charges $14/mo and watermarks on the free trial. Captions.ai is $9.99/mo with a monthly minute cap. ViralMint's caption tool is free in the desktop app — caption as many videos, as long as you want, with zero watermark.
Launch ViralMint and click Tools in the sidebar, then Captions. The page accepts mp4, mov, mkv, webm and most common video formats.
Drag your video onto the upload zone. ViralMint's Whisper engine starts transcribing immediately — word-level timestamps for the entire clip, processed locally.
Choose viral (yellow word highlight, 3 words at a time, Montserrat Bold), classic (full sentence, Arial), or bold (green highlight, 2 words, Impact). All three are presets — no font or color tuning required.
Click Render. ViralMint burns the captions in via FFmpeg using ASS subtitle format and writes the final mp4 to your output folder. The original video file is untouched.
Last updated May 2026
| Capability | Submagic | Captions.ai | Veed.io | ViralMint |
|---|---|---|---|---|
| Pricing | $14–29/mo | $9.99/mo | $24–59/mo | Free (desktop app) |
| Word-by-word animated captions | Yes | Yes | Yes | Yes — 3 presets (viral / classic / bold) |
| Watermark on free output | Yes (free trial) | N/A — paid only | Yes (free plan) | Never |
| Where transcription runs | Cloud | Cloud | Cloud | Locally (faster-whisper) |
| Monthly minute cap | Yes (plan-tiered) | Yes (60 min on cheapest) | Yes (plan-tiered) | None |
| Custom fonts / styling | Yes | Yes | Yes | Three curated viral presets (no manual tuning) |
| Open source | No | No | No | Yes |
highlighted column = clearer fit for that capability. Tied capabilities are left unmarked.
Yes. The Captions tool runs entirely in the ViralMint desktop app and uses local Whisper transcription plus FFmpeg for the caption burn — both free, no API key, no per-minute cost. Download the desktop app, drag a video into the Captions tool, and render. There's no monthly cap, no watermark on the output, and no upgrade prompt.
Three viral-tested presets out of the box: viral (yellow word-by-word highlight, Montserrat Bold 56pt, 3 words at a time), classic (full-sentence, Arial 42pt, bottom of frame, no per-word highlight), and bold (green word-by-word highlight, Impact 64pt, 2 words at a time). All three render via FFmpeg using ASS subtitle format, so timing is sample-accurate.
ViralMint uses faster-whisper running locally with int8 CPU quantization. On clean studio audio, word error rate is under 5% in English; for accented or noisy audio you may see occasional substitutions. Because the captions are word-level timed, you can edit any word in the generated subtitle file before re-rendering if you spot a mistake.
Yes — there's no length cap. Captioning runs proportional to clip length: a 5-minute clip takes about 30 seconds, a 60-minute podcast about 6–8 minutes on a typical laptop CPU. Local processing means you can caption hours of content overnight without metered cloud costs.
Yes — those are separate tools in the same app. The full pipeline can scout a trending video idea, generate a script, generate the voiceover (Edge TTS or paid gpt-4o-mini-tts), assemble the video against Pexels stock footage, then run the captioning step described here. You can use any single piece on its own, or chain them.
Whisper supports 100+ languages out of the box. The visible caption presets currently render Latin scripts cleanly; CJK and RTL scripts work but may need font swapping. Caption position, font and color are configurable per preset if you want to tune for a non-English audience.
Yes. The Captions tool is part of the ViralMint codebase that's open source on GitHub at github.com/openclaw-easy/ViralMint under the AGPL-3.0. The Whisper integration, ASS caption generator and FFmpeg burn-in are all open code you can read, fork and extend.
The AI Caption Generator ships inside the free ViralMint desktop app — no subscription, no watermark, no per-minute cap. Download once, use on as many videos as you want.
