What is Captions?

Captions is an AI-powered video creation studio founded in 2021 in New York City by Gaurav Misra and Dwight Churchill. It began life as a captioning app — hence the name — but has grown into a full creative suite that turns scripts into finished videos with AI avatars, editing, dubbing and voice cloning. It’s a serious operation: over 10 million creators, more than 3 million videos a month, and over $100 million raised from Sequoia, Andreessen Horowitz, Index Ventures and Kleiner Perkins.

In September 2025 the company rebranded its parent to Mirage and repositioned as a multimodal AI research lab for short-form video, though the consumer app you’ll actually use is still called Captions. The product is mobile-first and best on iPhone, where it has racked up 10 million-plus downloads. Where dedicated ad tools like Creatify are built around the campaign, Captions is built around the creator — the person making content who wants pro-level results without a studio or an editing degree.

Key features

Mirage Studio — the foundation-model engine

This is Captions’ technical showpiece. Rather than puppeting pre-recorded footage, Mirage Studio generates the entire clip — avatar, background and natural speech — from a foundation model. In practice that produces remarkably human motion: an actor can laugh mid-sentence, hesitate, then look away, none of it feeling scripted. It can even generate a complete video ad from nothing but an audio file. For emotional, testimonial-style reads, that range is genuinely hard to beat.

AI Creator, AI Avatars and AI Twin

The AI Creator feature builds a full video from a text script using a library of realistic avatars, so you can publish without ever being on camera — ideal for faceless channels or businesses pumping out marketing video. AI Twin goes further: upload a single selfie and Captions creates a personal avatar of you that you can re-dress, re-stage and put on camera saying anything you script.

A genuine editor, plus chat-based Cappy

Unlike single-purpose generators, Captions is a real editor. You can generate a talking-head clip, auto-caption it, drop in B-roll, adjust audio levels, rearrange scenes and export vertical — all in one workspace. The Cappy assistant, launched April 2026, lets you edit by texting instructions (“add music”, “tighten the intro”, “swap the look”), which is a strange but genuinely useful workflow.

Auto-Captions, dubbing and clean-up tools

The original strength still shines: fast, stylish, accurately-timed animated captions. Around it sits a suite of polish tools — AI Eye Contact to fix your gaze to camera, a one-tap Background Noise Remover, green-screen background removal, resize for any platform, and AI Dubbing across 28-plus languages with voice cloning and lip-sync.

The standout: the most natural motion, in the most complete app

Captions’ real edge is the combination — best-in-class natural motion from Mirage Studio, wrapped in the most complete creator toolkit in the category. Most rivals hand you a talking-head clip and stop; Captions takes you from script to avatar to captions to B-roll to dubbed, finished video without leaving the app, and the foundation-model output often clears the realism bar that puppeting-based tools can’t. For a solo creator or small team that wants one tool to do everything — and wants the talking avatar to actually emote — nothing here is as versatile.

Scorecard

Overall score: 7.9 / 10 — the average of the eight categories above.

Pricing

Captions has a free plan and several paid tiers, with the higher ones running on a credit system — the single most important thing to understand before committing.

The honest read: excellent value for editing and captions, expensive and unpredictable for AI avatars and ads. The free plan is genuinely generous — watermark-free edited exports are rare — and at $9.99 Pro is a steal for a creator who mainly wants captions and light editing. But the headline price is misleading if avatars are why you came: those run on Max and Scale, and true ad-grade Mirage Studio output sits all the way up at $399/month, where each second costs roughly 10 credits and iteration gets pricey fast. Worse, the credit system is hard to predict — it’s difficult to know what each AI task will cost, so heavy-use months can mean a much bigger bill than planned. Budget-conscious advertisers should map their real usage to a tier before subscribing.

Pros & cons

Honest weaknesses

The recurring complaint is the credit system. Across reviews, users find it genuinely hard to know how many credits a given AI task will cost, so monthly bills vary and budgeting becomes a guessing game — a real problem for anyone who needs predictable expenses. That unpredictability compounds at the top: the avatars most people come for aren’t on the cheap Pro tier at all, and the actually-ad-grade Mirage Studio output sits on a $399/month plan where roughly 10 credits per second means iteration adds up quickly. For a casual creator the value is excellent; for a business trying to forecast spend, it’s murkier than a flat per-video price.

Two other limits matter. First, despite the foundation-model firepower, Captions is creator-first, not ad-native — it lacks the URL-to-video automation, competitor intelligence and hook/body/CTA scaffolding that purpose-built ad tools like Creatify bake in, so performance marketers will do more manual lifting. Second, it’s firmly mobile-first: the iPhone app is excellent but the desktop and web versions feel less polished and Android users report more bugs, and the sheer number of features packed in can overwhelm newcomers. As with every avatar tool, the output can still tell on fast speech or big head turns, and AI-likeness raises the usual deepfake and disclosure considerations.

Who is Captions for?

Captions is the right tool for solo creators, faceless-channel operators and small teams who want one app to do everything — script-to-video, captions, editing, dubbing and a personal AI Twin — with the best natural motion in the category. If you’re producing a steady stream of short-form social content and value an all-in-one mobile workflow, the free and $9.99 tiers are superb value. It’s a weaker fit if you’re a performance marketer who wants an ad-native, URL-to-video pipeline — Creatify fits better there — if you need the most lifelike talking-head actors for paid social, where Arcads leads, or if you simply want the cheapest entry into AI UGC, where MakeUGC undercuts it. For broad general-purpose avatar video, HeyGen is the wider platform.

FAQ

AI UGC videoCaptionsMiragead creativeAI tool review