Hi Sumo-lings,

Matt here, founder of CaptionBolt.

It's been one week since we launched on AppSumo, and you've kept us busy in the best way. You've sent questions, reviews, emails and even your own videos, and I've read every one. Everything below started there.

There's a lot, so apart from the first one, I've kept each update to a line or two. You'll find the full details on our changelog: https://captionbolt.com/changelog

New: Magazine captions (every tier)

With Magazine, our goal was to make every frame of your video look like a magazine cover. The Captions panel now has 16 Magazine styles, from Cover Serif and Pull Quote to Front Page and Cluster. Each one sets a phrase as a designed page, with one large main word and the rest of the words laid out around it.

What makes it work:

• AI plans every page. It reads your transcript, splits it into pages, picks each page's main word and decides how strongly it should stand out, saving the biggest moments so they still feel big. Planning doesn't use processing minutes.

• AI reads every scene. When you upload, CaptionBolt follows where the speaker is throughout the video. Each page is placed to stay clear of their face, avoid bright or busy backgrounds and keep the main word easy to read.

• Text behind the speaker. Like the title on a real magazine cover, the main word can slip behind the speaker's head and follow its outline as they move. Styles marked Behind speaker do this only when the head is clearly visible, and only now and then, so it stays a special moment.

You stay in control. You can change any main word in Text Editor or Customize, and turn off Text behind the speaker in Customize. The preview, ASS download and exported video always match. Face-aware placement and Text behind the speaker work on new uploads, and they only look at where the speaker is, never who they are.

More of what you asked for

• Deep scan (from my last update): in Text Editor, recover captions hidden under songs, background music or fast speech. Included in every tier, with no extra processing minutes.

• Live style previews (also from my last update): hover over any style card in the Captions panel to watch it move before you pick it.

• Transparent caption layer (Tier 2 and Tier 3): export only your captions as a transparent ProRes 4444 .mov and place it over your video in your own editing app. It matches your video's resolution, so it lines up without resizing. Click Export caption layer next to Export video.

• Cantonese, Traditional and Simplified Chinese: choose the one you want when you upload, and the whole video stays in that script, including passages recovered by Deep scan. Cantonese uses written Chinese with Hong Kong Traditional characters, as Hong Kong subtitles usually do.

• Captions stay in sync when you cut words: delete words in Text Editor, and the caption no longer stays on screen while those words are spoken.

• No more captions at 0:00: if your video opens with silence or music, the first caption now waits for the first spoken word. This applies to new uploads.

Smaller improvements

• Five Talking styles (Talking Point, Card, Build, Focus and Spotlight) place short phrases at chest height for talking-head videos, with your highlighted words in an accent color.

• Five ready-made looks in the Headline Style tab: Clean, Label, Accent, Outline and Banner. Headlines also get a soft shadow, a fade in and out and better-balanced lines.

• Rename projects and transcripts: click the name at the top of the Editor, or choose Rename on the Dashboard or the Projects page. Copies are now named like "Interview (copy).mp4".

• Light mode: choose Light, Dark or System under Color mode in the account menu.

• An actions menu on Dashboard project cards, so you can edit, download, rename or remove a project without going to the Projects page.

• Captions, Headlines and Covers in Instrument Serif, Barlow Condensed or Bodoni Moda no longer render wider than planned.

Keep the suggestions coming. If something doesn't work the way you expected, or there's something you wish CaptionBolt could do, post in the Questions section here or email support@captionbolt.com. If it's about a specific video, include the project link. Real examples help us more than anything.

Thank you again for backing us this early. We'll keep listening, and we'll keep shipping.

Matt

CaptionBolt

CaptionBolt is purpose-built for one job: taking a short video you’ve already decided to publish and getting it social-ready as fast as possible. Here’s exactly what’s included and what isn’t — so you can buy with confidence.

✅ What CaptionBolt does

• Generates animated captions from the speech in your video — 338 styles, fully customizable (font, color, motion, word highlights)

• Deep Scan: recovers captions hidden under background music or fast speech

• Live style previews: hover to see a style animate before applying it

• Fix names and transcript mistakes in the Text Editor — no audio re-recording needed

• Resize to any social format: 9:16, 1:1, 4:5, or 16:9 — with crop, blurred background, or solid color fill

• Export 4K MP4s with no watermarks

• Export subtitles: TXT (all tiers), SRT + ASS (Tier 2+), VTT (Tier 3)

• Add a cover from 24 editable designs

• Add B-roll from Pexels, your own library, or AI suggestions

• Add your logo, progress bar, and saved brand presets

• Draft post copy for TikTok, Instagram Reels, and YouTube Shorts from your transcript

• 45 languages supported (AUTO mode for additional languages)

• REST API, job webhooks, and MCP server for AI agent automation (Tier 3)

• Works on mobile via your phone’s browser (no app download needed)

❌ What CaptionBolt does not do

• No AI clip-splitting — it doesn’t analyze long recordings to find viral moments; you bring it a clip you’ve already chosen

• No translation — captions stay in the language spoken; no cross-language subtitle generation

• No dubbing or voice synthesis — audio is untouched; only the captions and visuals change

• No long-term cloud storage — projects stay in your list for 14 days; download your finished MP4 and subtitle files right away

• No native phone app — mobile web works well, and a dedicated app is on our roadmap

• No custom SRT import — you can’t upload your own subtitle file as the transcription source (word-level timestamps are required for animated caption styles)

• No minute rollover — unused processing minutes reset each month

If what you need is automated clip-splitting from long recordings, translation, or dubbing — you’ll need a different tool for those steps. But if you have a clip ready and just need to finish it, the workflow here is fast and direct.

Questions? Post below or email support@captionbolt.com.

— Matt, Founder of CaptionBolt

Hi Sumo-lings,

Matt here, founder of CaptionBolt.

Thank you. Since our AppSumo launch, you've put CaptionBolt to work on your own videos and told us honestly what works and what doesn't. For a small, bootstrapped team like ours, that kind of feedback is everything. I read every question, review, and suggestion, and both updates below came straight from them.

1) Deep scan: recover captions hidden under music

Some of you pointed out that in videos with songs, heavy background music, or very fast speech, a stretch of captions could go missing, or a single word would stay on screen far too long. You were right, and I'm sorry if that happened on a video you care about.

This turned out to be one of the toughest problems in captioning. Music and fast vocals are the hardest audio there is for speech recognition, and these gaps appear without any warning, so they're easy to miss. Over the past few days, we studied dozens of real videos and tested the leading speech recognition AI side by side. None of them handle this perfectly yet. It's a limit of where AI is today.

So rather than wait for AI to catch up, we built Deep scan. It finds those spots for you and gives each one a focused, deeper analysis to recover the vocals and lyrics hidden under the music. In our testing, it brought back most of what was missing.

To use it, open a project, go to Text Editor, and click Deep scan. Play each recovered passage against the original audio, keep the ones you want, and add them in one step you can undo.

A few honest notes:

• Clear voices give the best results. Pure music, humming, or voices buried under noise may not be recoverable.

• Like any AI, it can occasionally mishear a lyric, and recovered lines may need a quick touch-up, such as capitalization or punctuation. That's why you can play every passage and choose what to add.

• It's included in every AppSumo tier and doesn't deduct extra processing minutes.

• It works on projects uploaded in the last 14 days, including ones uploaded before this update. For a completed project, click Edit again to make an editable copy, then run Deep scan there.

We'll keep improving this as speech recognition AI gets better.

2) Live style previews: see a style move before you pick it

Others asked for an easier way to see how a caption style actually moves before applying it. Now, in the Captions panel, just hover over any style card and it plays a short loop of that style's animation. You can compare styles at a glance instead of trying each one on your video.

Keep the suggestions coming. If something doesn't work the way you expected, or there's something you wish CaptionBolt could do, post in the Questions section here or email support@captionbolt.com. If it's about a specific video, include the project link. Real examples help us more than anything.

Thank you again for backing us this early. We'll keep listening, and we'll keep shipping.

Matt
CaptionBolt