Captions

  • Captions — Mobile-First AI Short-Form Video Studio: Auto Captions, Eye Contact Fix, AI Dubbing, AI Twins & Agentic Edit (Mirage-backed, Active)

    Website

    Field Value
    Official Website https://captions.ai
    Parent Company Mirage (rebranded from Captions Inc., Sep 2025; NYC, founded 2021 by Gaurav Misra + Dwight Churchill; 75M GC growth Mar 2026; 20M+ users, 250M+ videos)
    Status Active, independent, not shut down (product name “Captions” retained under Mirage umbrella)
    Category ai-productivity-data
    Pipeline Stage none

    Title

    Field Value
    Product Captions
    Descriptive Title Captions — Mobile-First AI Short-Form Video Studio: Auto Captions, Eye Contact Fix, AI Dubbing, AI Twins & Agentic Edit (Mirage-backed, Active)
    One-line Positioning Talking-head-to-Reels in one pass: teleprompter record → AI captions (100+ langs) + eye-contact correction + filler removal + voice-clone dub with lip-sync → vertical export; generative layer (AI Twin/Actor/Chat-to-Edit) on credit meter.

    Features

    Field Value
    Captioning Auto speech-to-text, 100+ languages, word-level animated styles (karaoke/typewriter/bounce), emoji timing, editable
    Recording Built-in teleprompter with AI Script, Captions Camera, gaze kept to lens via post-hoc Eye Contact
    Audio AI Filler-word/pause removal, AI Denoise, AI Voice Clone, AI Voiceover, AI Sound Effects, AI Censor
    Visual AI Auto reframe 16:9→9:16, zoom/keyframe, smart cutaways, B-roll insert, background removal
    Dubbing Voice-clone translate 29-30+ langs with lip-sync (AI Lipdub)
    Generative (credit) AI Edit (style presets), Chat-to-Edit, AI Creator (text→footage), AI Twin (selfie avatar, up to 30 on Max), AI Actors (Mirage model), AI Ads (URL→UGC), AI Shorts (long→clip), Reddit-to-Reel, AI Image/Music/Sounds
    Cloud Glue Own Mirage foundation model (pacing/framing/attention), proprietary ASR + lip-sync; no GCP/AWS coupling
    Export 9:16 + 16:9, resize, no watermark on paid, transcript download

    Pricing (2026-08 USD/mo, monthly; annual cheaper)

    Field Value
    Free $0 — 60-200 lifetime credits (no refresh), trim/zoom/transitions/captions camera/basic captions, 1 template (Eclipse), no generative AI, no watermark on allowed exports
    Lite (Android only) $4.99 — manual edit + translate/dub/eye-contact/denoise/copy-transcript
    Basic (formerly Pro) $9.99 — 200 credits/mo, no watermark, Eye Contact/Denoise/Censor/Voiceover/Trim, Caption&Dub, Reddit-to-Video, teleprompter
    Max $24.99 — 500 credits/mo, Chat-to-Edit, AI Edit styles, Mirage actors, up to 30 AI Twins, concurrent gen (2)
    Scale $69.99 — 1,400 credits/mo, fastest models, 3× AI usage headroom
    Scale 2x $139.99 — 2,800 credits
    Scale 4x $279.99 — 5,600 credits
    Enterprise Custom — bulk credit, seats, training-data exclusion, SLA
    Credit Rule Monthly refresh; Free credits lifetime non-refreshing; prompt-to-video ~16 cr/6s, AI Edit 10-40 cr, GenAI video 1-20 cr; starting_paid=$9.99; free_tier=yellow-60-200-lifetime-credits

    Reviews

    Field Value
    Apple App Store ~4.6-4.7/5 (regional slices, 180-200+ ratings) — caption styles/translate praised; trial-charged-early, mid-sub feature moved to Mirage paywall, audio drift griped
    Trustpilot 4.0/5 (~253) per RecaTools aggregate; AIToolTier-style scrape 1.6/5 — support unresponsive, billing/refund, avatar pull mid-sub dominate negatives
    Capterra 5.0/5 (12-21 reviews, too thin to weight); ease 4.7 / value 4.2 sub-scores
    G2 Not enough reviews to rate (excluded)
    Editorial ToolCurrent 8.0/10; RecaTools mixed — captioning best-in-class, export audio-sync + credit burn + support drag
    Praise Caption styling best on Reels/TikTok, eye-contact fix real, dub+lipsync standout, teleprompter flow, 1-pass short-form
    Gripes Audio desync on export, credit burn on Twin/Dub, App Store trial charged early, 2025-26 AI Twin/Actor shifted to Mirage app paywall mid-cycle, crash on long clips, support slow on refund-when-credits-consumed

    Review Sources Breakdown

    Field Value
    Apple App Store (apps.apple.com) 4.6-4.7/5, ~180-200 ratings, installer-base signal, consumer trial/billing/feature-move complaints
    Trustpilot (trustpilot.com) 4.0/5 (~253) per RecaTools; scraped subset 1.6/5 dominated by support/billing/refund, not feature-quality
    Capterra 5.0/5 but 12-21 reviews → statistically not material
    G2 “Not enough reviews” → excluded from any average
    Editorial (ToolCurrent / RecaTools / ViralClips) 8.0/10, 6.5/10-style single-editor scores, feature confirmation only
    Vendor blog Funding/user-count only, not ratings

    Access

    Field Value
    Mobile iOS app (App Store 1541407007), Android app
    Desktop macOS app, Web studio (captions.ai/app)
    CLI None
    API REST API on Scale/Enterprise (programmatic video gen), Mirage API separate; no public MCP, no self-host
    Integrations Direct export TikTok/Reels/Shorts/YouTube; no Figma/Discord/ComfyUI; Zapier not first-class
    Login Apple/Google/email; consumer generative features migrating to Mirage app umbrella

    Best For

    Field Value
    Audience 1 Talking-head short-form creators (coaches, influencers, podcasters) on TikTok/Reels/Shorts
    Audience 2 Solopreneurs wanting teleprompter→captioned reel in 5 min, no timeline
    Audience 3 Small brands making UGC ad variants from product URL (AI Ads)
    Audience 4 Agencies doing multilingual dub+lipsync for 29+ markets at low cost

    Not For

    Field Value
    Exclude 1 Long-form/documentary/multi-cam editors (→ Premiere/CapCut/Descript)
    Exclude 2 Timeline-level control freaks (caption font/size/color limited vs NLE)
    Exclude 3 MCP/self-host/own-LoRA users (→ Runway/Kling/Flux local)
    Exclude 4 Budget solo hitting credit wall fast (Twin/Dub eat credits, $9.99 too thin)
    Exclude 5 Enterprise needing SOC2/HIPAA paperwork (Captions app not compliant; Mirage Enterprise side only)

    Competitors

    Field Value
    CapCut ByteDance free editor, massive distribution, weaker caption taste/lipsync, no Twin
    Descript Audio-first doc-style edit, strong overdub, weaker vertical caption aesthetics
    Opus Clip Long→short clip specialist, no caption styling/dub/twin depth
    HeyGen AI avatar presenter web-first, no mobile caption pipeline, $29+ start
    Synthesia Corporate training avatars, no short-form social captioning
    Runway Generative video models, not talking-head caption workflow
    Pictory Text/article→video, no eye-contact/lipsync
    Kapwing Browser editor w/ AI, broader but shallower on lipsync/dub

     

Do Not Sell or Share My Personal Information Cookie Settings