AI video editor that does talking-head magic: virtual studio, eye contact correction, AI avatars, multilingual dubbing. Mobile-first, surprisingly powerful.
Video
Captions is an iOS-first (also on Android and web) AI video editor built for one job: turning a phone-recorded talking-head clip into a captioned, B-rolled, filler-free short-form video without opening a timeline. Point the camera at yourself, hit record, feed the raw take to Captions, and the app returns a polished vertical export with styled subtitles, cutaway B-roll, background music, and — if you want it — an AI-generated avatar reading a script you never filmed. The company claims 20 million users and supports captions in 100+ languages, which places it in serious scale territory for a creator app.
The honest pitch: Captions is for the solo creator or small brand who ships one or two short-form videos a day, films them on a phone, and wants the edit collapsed to a chat prompt. It's not for editors who want frame-accurate control, and it's not for teams collaborating on longer-form podcast or explainer content — Descript owns that lane. If your workflow is "record on phone, post to TikTok, Reels or Shorts within the hour," Captions was designed for you.
Against the market, Captions sits between two competitor clusters. On one side you have CapCut and Adobe Premiere Rush — free or near-free mobile editors with traditional timelines and shallow AI. On the other side you have Descript, Opus Clip, and Vizard — AI-heavy tools that lean toward repurposing longer footage (podcasts, webinars, YouTube uploads) into clips. Captions plants itself at the intersection: mobile-first capture, AI-first editing, generative avatars for creators who don't want to film every take themselves. The trade-off, and it's a real one, is that Captions is meaningfully more expensive than any of those alternatives once you cross into paid tiers.
Chat-Based AI Edit. The headline feature. Upload a raw clip or record in-app, then type instructions in plain English — "cut the pauses, add B-roll every 5 seconds, use the Prism Pro caption style" — and Captions executes the edit. It's the same conceptual pattern as Descript's Underlord, but built for vertical video and creator workflows rather than podcast timelines. Quality varies with prompt specificity; expect to babysit the first few edits until you learn its idiom.
AI Twin (Custom Avatar). Record a short training clip of yourself and Captions builds a digital twin that can read any script in your voice and face. HeyGen and Synthesia offer similar avatars, but Captions bundles the twin into the same app you're already editing in, and it's tuned specifically for creator-length shorts rather than corporate training videos. Realism is good on head-and-shoulders framing, weaker on hand gestures or full-body shots.
Auto Captions with 100+ Templates. Word-by-word styled subtitles in 100+ languages, with animated templates named things like Bloom, Prism Pro, Chalk, and Y2K. Font, color, position, and highlight behavior are all editable. This is the feature Captions built its reputation on and it remains best-in-class for aesthetic variety — most competitors ship five or ten caption styles and stop.
Eye Contact Correction. Fixes the "reading from teleprompter or notes" gaze problem by warping your pupils to look at the lens. Useful for creators reading scripts off-screen; noticeable and slightly uncanny if overused or if the source footage has extreme angles. Descript ships a similar feature; Captions' version tends to be more aggressive.
Pause Trimming and Filler Removal. Detects "um," "uh," dead air, and long pauses and cuts them automatically. Standard AI-editor table stakes at this point — Descript, Opus Clip, and Vizard all do this — but Captions' implementation is fast and rarely mangles cadence.
B-Roll, Music, and Sound Effect Generation. Captions pulls or generates cutaway visuals and adds royalty-free music beds keyed to the topic of your monologue. Credit-metered on paid plans; expect to spend real credits on any video that leans heavily on B-roll.
Voice Cloning and Translation. Clone your voice and translate a video into another language while keeping your voice and lip-sync intact. This is the feature that justifies the price jump to Scale tiers for creators publishing multilingual content. Quality on major European languages is strong; less consistent on tonal languages.
Mobile-First Capture Loop. The one area Descript and Opus Clip genuinely can't match: Captions was designed phone-first. You can film, edit, and post from the same app without ever touching a laptop. Web and Android exist, but iOS is the flagship — the pricing page even discloses it quotes "iOS plans only."
Captions runs a credit-metered model with a free tier and four paid tiers on iOS. All AI edits, B-roll generations, avatar renders, and translations draw from a monthly credit pool that resets on your billing anniversary.
| Tier | Price | Credits / month | Best for |
|---|---|---|---|
| Free | $0 | Limited, no generative AI | Trying auto-captions before paying |
| Max | $24.99/mo | 500 | Solo creator, 1-2 videos/day |
| Scale 1x | $69.99/mo | 1,400 | Daily poster with heavy B-roll |
| Scale 2x | $139.99/mo | 2,800 | Small agency or multi-account operator |
| Scale 4x | $279.99/mo | 5,600 | High-volume shop or creator team |
| Enterprise | Custom | Bulk credits | Brand deployments, training-data exclusion |
Two things to notice. First, there is no cheap paid tier — the jump from free to Max is $24.99/month, and the jump from Max to genuinely-unlimited-feeling usage is Scale 1x at $69.99. Second, the credit system means power features (avatars, translation, generative B-roll) burn credits fast; heavy users routinely hit the Max ceiling mid-month and either upgrade or ration.
For comparison anchors: Descript Creator is $24/mo on annual billing ($35 monthly) with 30 hours of media and 800 AI credits — a fundamentally different pool sized for podcast-length editing. Opus Clip Starter is $15/mo, Pro $29/mo for watermark-free repurposing from long videos. CapCut Pro is roughly $9.99/mo ([verify current pricing] — CapCut varies by geography and promo) and gives you a full timeline editor with lighter AI. Vizard Creator [verify pricing — page currently shows promotional 50% off with regular rate obscured] targets repurposing at a lower entry point. Captions is priced at the top of the creator-tool bracket, and the pitch has to justify that gap on avatar quality, caption aesthetics, and chat-based editing speed.
Pros
Cons
Descript — Creator $24/mo annual ($35 monthly). Pick Descript if you edit podcasts, YouTube long-form, or anything transcript-driven. Its text-based editor is unmatched for spoken-word content and 30 hours of monthly media is genuinely generous. Pick Captions instead if your output is vertical shorts filmed on a phone and avatar generation matters to you.
Opus Clip — Starter $15/mo, Pro $29/mo. Pick Opus Clip if your workflow starts from a long YouTube video, podcast episode, or webinar that you want auto-clipped into shorts with virality scoring. It's cheaper and more repurposing-focused. Pick Captions if you shoot short-form natively and want caption aesthetics and avatar features Opus doesn't offer.
CapCut — Free tier is generous, Pro roughly $9.99/mo [verify]. Pick CapCut if you want a full timeline editor, are cost-sensitive, and are comfortable doing more work manually. It's the default for a reason — free, powerful, cross-platform. Pick Captions if you'd rather trade money for time and let AI make editing decisions for you.
Vizard — Free tier plus Creator and Business [verify current pricing]. Pick Vizard if you're repurposing long-form to shorts and want scheduled posting to multiple social accounts built in. Pick Captions for higher caption template quality and AI Twin avatars.
Adobe Premiere Rush — $9.99/mo standalone, bundled in Creative Cloud. Pick Rush if you're already in the Adobe ecosystem and want deeper manual controls with cross-device projects. Pick Captions if AI-first editing and generative avatars are the point of the tool.
Buy Captions Max ($24.99/mo) if you are a solo creator who ships one or two talking-head shorts a day on iOS, wants best-in-class caption styling, and values speed over price. Upgrade to Scale 1x ($69.99/mo) only when you're consistently hitting the Max credit ceiling and generating B-roll or avatars on most videos. Skip Captions entirely if you're editing long-form podcasts (use Descript), repurposing YouTube videos into clips (use Opus Clip or Vizard), or you're cost-sensitive and comfortable with a traditional timeline (use CapCut). The tool is real and the quality is real, but the pricing assumes your time is worth more than the $10-$25/month gap between Captions and every named competitor — if that's true for you, it's worth it; if not, one of the alternatives will do 80% of the job for 40% of the price.
Q: Is Captions free? A: Captions offers a free tier with basic editing, media library assets, captions in 100+ languages, and one caption template — but no generative AI features. To unlock chat-based editing, AI avatars, and B-roll generation you need Max at $24.99/month or higher.
Q: How does Captions compare to Descript? A: Captions is built for phone-shot vertical shorts; Descript is built for transcript-driven long-form (podcasts, YouTube videos, webinars). Descript Creator is $24/mo annual with 30 hours of media, Captions Max is $24.99/mo with 500 credits. Different jobs, near-identical price point.
Q: Does Captions work on Android and desktop? A: Yes — iOS, Android, and web are all supported. But the pricing page explicitly notes plans reflect "iOS plans only," and iOS remains the primary development target. Android and web users should expect a slight lag on feature parity.
Q: What are Captions credits used for? A: Credits are consumed by generative AI actions — AI Edit runs, B-roll generation, avatar renders, translations, and voice-clone outputs. Basic captioning and trim edits don't burn credits. Max includes 500/month; Scale 1x jumps to 1,400.
Q: How good is the AI Twin avatar feature? A: Realism is strong on head-and-shoulders framing with steady lighting, weaker on full-body shots or animated hand gestures. Comparable in quality to HeyGen and Synthesia for talking-head use cases, but Captions bundles it into the same editing app rather than a separate corporate video platform.
Q: Can Captions translate videos into other languages? A: Yes — Captions supports voice-cloned translation with lip-sync into 100+ languages. Quality is strong on major European languages (Spanish, French, German, Portuguese) and less consistent on tonal languages like Mandarin or Vietnamese. This feature draws heavily on your credit balance.
Q: Is Captions worth it over CapCut? A: If your priority is doing less manual editing and having AI make the cuts, transitions, and B-roll decisions for you, yes. If your priority is cost and you're comfortable with a traditional timeline, CapCut Pro (roughly $9.99/mo [verify]) does most of the same edits manually for a fraction of the price.
Q: Does Captions add a watermark on the free tier? A: The free plan is limited and free-tier exports carry Captions branding based on the app's current terms. Paid tiers (Max and above) export watermark-free.