- Starting price
- First 4 hours free · paid plans from $16/month
- Free tier
- Yes
- Platforms
- Web · Mac · iPhone · iPad · Android
- Developer
- Hunt Creative House / James Hunt
- Launched
- 2023 platform · 2024 mobile app
- Updated
- Aug 9, 2026
The verdict
CaptionKit is a credible church-first accessibility service with a low-risk four-hour trial, browser-based broadcast, QR access, free attendee apps, text-to-speech, display layouts and useful production integrations. Its modular $12 license plus caption-hour and translation pricing can be economical for a weekly service, though costs rise with usage. Churches should run native-speaker and Deaf/hard-of-hearing pilots, define human backup for critical moments, and ask for a clearer service-specific privacy and data-retention statement before sending pastoral audio every week.
Try CaptionKit ↗Opens captionkit.com
CaptionKit turns live church audio into same-language captions and translated text that people can read on phones, tablets, computers or room displays. A church streams sound into the browser or Mac app, chooses the spoken language and any translations, then shares a public link or QR code. Attendees can open the result without installing anything. The free iOS and Android apps add adjustable text, dark or light display, church search and text-to-speech so translated words can be heard through headphones or compatible Bluetooth hearing aids.
The platform is built for a specific accessibility gap. Professional interpreters remain the strongest option where available, but many congregations cannot provide an interpreter for every language and gathering. Generic auto-caption tools often misunderstand biblical names or cannot feed the exact screens used in a sanctuary. CaptionKit supplies boosted and blocked keywords, profanity filtering, lower-third, full-screen and styled display layouts, history and viewer analytics, and links for ProPresenter, OBS, vMix, FreeShow, WorshipTools Presenter, Proclaim and production automation. Official documentation currently says more than 65 spoken inputs and more than 85 translation outputs; an older support summary still says 55 and 80, so teams should use the live language list rather than a directory number.
Pricing is refreshingly visible. A new church receives four hours of captions and translations without a credit card. Paid configuration is modular: the site shows a $12 monthly church license, four caption hours for $4 and up to two simultaneous translations for another $4, producing a $20 example monthly plan. It also says plans begin at $16, are month-to-month and can be changed. That structure is attractive for a single weekly sermon but requires calculation for worship, announcements, classes, multiple campuses, conferences and long services.
This researched review uses the current product and pricing page, Help Center, supported-language record, privacy policy, Apple and Google listings and release history. We did not stream a service, measure latency, score word error rates, assess every dialect, connect a soundboard, test presentation software, validate text-to-speech in hearing aids, inspect transcripts, confirm deletion, run a security review or interview congregants. Apple’s four-rating average is 4.0 and Android shows more than one thousand downloads. Those are useful identity signals, not enough evidence to predict performance in a particular room.
✓ The good
- No-card pilot - four free hours let a church test its real pastor, sound system, language and network before paying
- Attendee flexibility - captions open from a QR link on any modern browser, while free apps add text-to-speech and display controls
- Broad language coverage - the current site advertises 65+ input languages and 85+ translation outputs
- Production-ready displays - lower thirds, full screen and styled layouts can feed dedicated monitors or livestream compositions
- Church vocabulary controls - boosted or blocked keywords help with names, places and recurring theological terminology
- Useful integrations - official guides cover ProPresenter, OBS, vMix, FreeShow, WorshipTools Presenter, Proclaim, Companion and cURL
- Transparent modular pricing - the site exposes license, caption-hour and translation components without requiring a sales call
- Unlimited team and streaming-device claim - staff access is not positioned as a per-seat surcharge
✗ Watch out
- Automated accuracy varies - accents, room echo, music, overlapping speech and code-switching can still produce consequential errors
- Usage can exceed the headline price - full services, classes, multiple campuses and several output languages consume more than four hours
- Thin rating evidence - four US Apple ratings and a modest Android install band do not establish reliability at scale
- Generic privacy policy - the Australian policy does not clearly explain audio, transcripts, translation vendors, retention or church attendee analytics
- Store labels conflict - Apple lists non-linked usage and diagnostics while Google says the app collects no data
- No declared Apple accessibility - the product is accessibility-oriented but Apple shows no developer declarations for supported features
- Internet dependency - live cloud transcription and translation require stable upstream audio and connectivity
- Human interpretation remains necessary in some settings - pastoral care, sacraments, emergencies and nuanced multilingual ministry exceed automated text
Best for
- Churches beginning a Deaf and hard-of-hearing accessibility ministry with limited interpreter availability
- Congregations serving one or two recurring language communities during a weekly sermon
- Production teams that need live captions in ProPresenter, OBS, vMix, FreeShow, Proclaim or a dedicated display
- Attendees who prefer translated text or text-to-speech through their own phone and headphones
- Churches willing to run monitored pilots and maintain a human correction or fallback process
Avoid if
- A certified interpreter or legally assured accommodation is required and automation alone would be insufficient
- Your venue lacks stable internet, a clean soundboard feed or a volunteer able to monitor the session
- Many simultaneous translated languages and long weekly schedules make metered hours uneconomical
- You need a contractual answer about transcript retention, subprocessors, data location and deletion before any trial
- The congregation expects translation to preserve tone, cultural nuance and theological precision without human review
What CaptionKit is
CaptionKit has two connected sides. The church-facing control layer receives a live audio feed, runs transcription and translation, applies vocabulary settings and publishes browser or presentation outputs. The attendee-facing layer displays the selected stream through a public web link or mobile app. That separation keeps installation optional: guests can scan and read immediately, while regular users can choose the richer native experience.
It is not prerecorded video subtitling software, an interpreter marketplace or a general church-management system. It handles speech occurring now. Session history and analytics can help a team understand use, and downloadable transcripts may support later workflows, but the primary promise is inclusion during the gathering. Because the product transmits live service audio to cloud systems, accessibility planning and data governance belong in the same adoption conversation.
Why church production teams choose CaptionKit
The differentiator is not speech recognition by itself; modern phones and streaming platforms already provide some captions. CaptionKit packages recognition into church operations. Volunteers can boost the unusual name in this week’s sermon, present captions as a lower third in the existing video workflow, display a full-screen feed for the room and offer translated text on personal devices from one session. cURL, Companion and ProPresenter controls reduce the number of manual steps during a service.
Its modular pricing also suits churches that cannot justify an enterprise accessibility contract. Four hours without a card makes a serious pilot possible, and the sample $20 plan roughly maps to one captioned sermon per week. The discipline is to test the entire path with the people it is intended to serve. A clean demo spoken into a laptop is not representative of a reverberant sanctuary, soft reader, accented guest, music transition or live theological vocabulary. Invite native speakers and Deaf or hard-of-hearing congregants to judge usefulness, not merely technical staff.
Live captions and translation: broad reach with variable accuracy
The church selects a spoken input and can offer multiple translated outputs. The current marketing page says CaptionKit understands sermons in more than 65 languages and translates into more than 85; the detailed May 2026 support article enumerates inputs, dialects and translation-only additions. Attendees choose their language from the shared service, and same-language captions remain useful for hearing access, noisy rooms, attention and unfamiliar accents. The mobile app can read translated text aloud through headphones or compatible hearing devices.
Language count is not quality assurance. Speech recognition degrades with poor gain staging, distance, echo, clipping, laughter, multiple voices and music. Translation adds another interpretive layer and may flatten idiom or select the wrong theological sense. Run representative recordings and live services, seed names and terms, place a bilingual monitor in the loop and publish a simple disclaimer. Do not let machine output become the sole channel for safety instructions, pastoral counseling, vows or another moment where a mistranslation could materially harm someone.
Displays, links and production control: one stream in several places
A public caption link or QR code works in modern browsers, reducing guest friction. In-room or livestream outputs include scrolling lower thirds, full-screen scrolling captions and styled non-scrolling layouts. Official guides document use with common church production products, and automation can call cURL endpoints, Bitfocus Companion or ProPresenter actions. History records past sessions and viewer analytics show how streams were consumed. Team and streaming-device access is advertised without a numerical seat limit.
A good deployment treats captions as a system, not a URL. Decide who starts the stream, verifies the selected languages, checks that the public link is discoverable, watches for stalled captions and closes the session. Keep the QR code high-contrast and available before speaking begins; provide a short human explanation for first-time users; and maintain a fallback such as front-row seating, printed material or an interpreter. Record what happens when the Mac, browser, audio interface or internet connection fails mid-service.
Privacy and governance: the weakest part of the public documentation
CaptionKit’s privacy policy invokes the Australian Privacy Principles and describes collection of names, addresses, email addresses, phone numbers and information obtained through correspondence, public sources and third parties. It allows consented or legally required disclosure, promises reasonable security and says client files may be kept for at least seven years. Apple declares non-linked product interaction, crash and performance data for analytics and functionality. Google’s mobile listing says no data is collected or shared.
The policy does not clearly state whether live audio is retained, whether transcripts are stored in session history, which speech and translation processors receive them, where data is hosted, how long attendee analytics persist or how a church deletes a session. Those are central questions when sermons include prayer requests, testimony or identifiable pastoral information. Before ongoing adoption, request a data-processing agreement or written answers covering audio, transcripts, subprocessors, location, encryption, retention, deletion, account access and incident response. The store-label mismatch should be reconciled as part of that answer.
Pricing
Free trial
4 hours free
Four hours of live captions and translations with no credit card required; suitable for a real-service pilot.
Church license
$12/month
Base monthly access before caption hours and simultaneous translation capacity are added.
Example weekly plan
$20/month
$12 church license plus $4 for up to four caption hours and $4 for up to two translations at once.
Configured plans
From $16/month
Caption hours and translation capacity can be adjusted monthly; calculate the live configuration on the official pricing tool.
CaptionKit begins with four caption and translation hours at no charge and requires no payment card. Use those hours on the exact sound system, network, languages and volunteer workflow intended for launch. A quiet office test is useful for setup but should not consume the only serious evaluation.
The public configurator separates a $12 monthly church license from usage. Its example adds four caption hours for $4 and two simultaneous translations for $4, totaling $20 per month. The headline “from $16” represents a smaller configuration, not unlimited captioning. Plans are billed monthly and can be changed without a long contract.
Build a realistic hours model. Count the full time between starting and ending transcription, not only sermon length. Add rehearsals, announcements, prayers, classes, youth meetings, holiday services, multiple campuses and the staff time spent monitoring. Ask how overruns are billed and whether simultaneous streams use hours in parallel before depending on a budget estimate.
The attendee web experience and mobile downloads are free, but the church must subscribe after the trial to broadcast. No mobile purchase unlocks service independently. This avoids charging the person who needs accommodation, a sound design choice, while placing procurement and governance where they belong—with the broadcasting organization.
Where CaptionKit falls behind
Privacy specificity - a generic policy is not enough for live audio, saved transcripts, attendee analytics and third-party AI processing.
Independent accuracy evidence - testimonials are encouraging, but public per-language word-error and translation-quality benchmarks are absent.
Rating depth - four Apple ratings and 1K+ Android downloads remain a small public footprint for an operational service.
Offline resilience - cloud recognition cannot continue through an internet outage, so production teams need a documented fallback.
Accessibility documentation - Apple lists no declared support even though accessibility is the product’s central purpose.
Cost predictability at scale - modular hours are economical for light use but require careful modeling for long or multi-campus schedules.
CaptionKit vs. Kaleo AI vs. human interpreters
CaptionKit emphasizes flexible production displays, church vocabulary controls and modular low-volume pricing. Its current site advertises 65+ spoken inputs and 85+ outputs, with a $20 example for four hours and two translations. Kaleo AI advertises custom training for theological terminology, 150+ translation outputs and simpler fixed tiers: $24 for five hours and two languages, $49 for ten hours and four, or $149 unlimited. Both offer browser access, QR codes and mobile spoken translation; they are separate products from separate companies despite similar attendee app descriptions.
A professional interpreter offers context, tone, correction, cultural mediation and bidirectional conversation that automated captions do not. Interpreters cost more and may be unavailable for every language or weekly gathering, while AI can extend baseline access rapidly. The strongest model can be layered: reliable captions for broad ongoing reach, qualified interpreters for recurring communities and critical services, and bilingual reviewers who improve vocabulary and catch errors.
Choose CaptionKit when presentation integration, a no-card trial and granular small-church pricing matter. Evaluate Kaleo when the language catalog or predictable fixed tiers fit better. Choose human interpretation when nuance, trust, legal accommodation or interactive communication is central. Never decide solely from the number of languages on a pricing page; a tested high-quality output in the congregation’s actual language matters more than a theoretical list.
The bottom line
CaptionKit makes a difficult accessibility project unusually approachable. A church can try four real hours without a card, send captions to the room and livestream, let visitors scan a link, and offer translated audio through personal headphones. The product team has also documented production integrations better than many young competitors. The public privacy explanation is the conspicuous gap, and machine accuracy will always depend on audio, language and human oversight. Pilot with the congregation, measure comprehension rather than novelty, get written data answers and budget the full service calendar. Under those conditions, CaptionKit can be a valuable accessibility layer—never a claim that the language barrier has been solved once and for all.
Alternatives to CaptionKit
Kaleo AIChurch-focused live translation into 150+ outputs with fixed five-hour, ten-hour and unlimited plans.
ProPresenterA full church presentation platform that can display CaptionKit output but is not the translation service itself.
FreeShowOpen-source presentation software with an official CaptionKit connection guide.
Faithlife ProclaimCollaborative church presentation software with its own ecosystem and a CaptionKit display path.Microsoft TranslatorA general-purpose translation option with broader use cases but less church production specialization.Frequently asked questions
How much does CaptionKit cost?
The first four hours are free without a credit card. The site shows a $12 monthly church license, $4 for up to four caption hours and $4 for up to two simultaneous translations, producing a $20 example plan. It advertises paid configurations from $16 per month and month-to-month changes.
Do church members need the CaptionKit app?
No. Attendees can scan a QR code or open the church’s public link in a modern browser. The free iOS and Android apps add native church search, adjustable text, light and dark modes, and text-to-speech for translated audio through headphones or compatible hearing aids.
How many languages does CaptionKit support?
The current home page says more than 65 spoken input languages and more than 85 translation outputs. A May 2026 support article gives the current detailed list. Older summaries still mention 55 and 80, so confirm the exact input dialect and desired output in the live documentation.
Does CaptionKit work with ProPresenter or OBS?
Yes. Official materials document lower-third, full-screen and styled browser displays and connections for ProPresenter, OBS, vMix, FreeShow, WorshipTools Presenter and Proclaim. It also advertises cURL, Bitfocus Companion and ProPresenter automation controls.
Can CaptionKit replace a sign-language or spoken-language interpreter?
Not in every setting. It can widen baseline access when interpreters are unavailable, but automated text can miss tone, culture, names and theological nuance. Use qualified interpreters for legally or pastorally critical communication and involve the people served in evaluating the system.
Does CaptionKit save church audio or transcripts?
The public privacy policy does not answer that with enough specificity. It mentions personal information, third parties, security and seven-year client-file retention, while product materials advertise session history. Ask CaptionKit directly about audio, transcripts, subprocessors, hosting, retention, deletion and attendee analytics before recurring use.
Is CaptionKit the same product as Kaleo AI?
No. CaptionKit is operated by James Hunt/Hunt Creative House with captionkit.com and app bundle io.captionkit.app. Kaleo AI is operated by EID Systems Inc. with withkaleo.com and bundle ai.getkaleo.mobile. Their attendee apps describe similar accessibility workflows, but their companies, pricing, language claims, policies and store records are distinct.
Sources & further reading
More Worship Apps
Kaleo AIKaleo AI streams church services into live captions and 150+ translated languages, with theology-aware vocabulary and simple fixed plans—but human review and stronger data-governance answers remain essential.
AsaphAn ambitious AI workspace for sermon-aligned setlists, song analysis and local feedback—valuable as a planning assistant, risky if its theology scores become a substitute for pastoral judgment.
Nepali Christian Lyrics & ChordsA rare, actively maintained songbook built around the real needs of Nepali-speaking worship teams—but privacy and song-rights documentation lag far behind the app.
OnStageA modern all-in-one workspace for worship charts, setlists, rehearsals, schedules, chat, and live performance—but conflicting prices and missing legal pages make due diligence essential.
Unceasing WorshipIHOPKC’s niche streaming service preserves spontaneous prayer-room worship, intercession, live archives, and 76 legacy albums—valuable for listeners who want that exact charismatic worship tradition, but not a general Christian music replacement.
MultiTracks.comThe officially licensed stem library and Playback session player behind most touring worship rigs and large-church Sunday mornings - and the closest thing the industry has to a default.