Back to Just Captions

Privacy Policy

Last updated September 27, 2026

Just Captions is designed for adding captions to videos on your device. The app does not require an account and does not include advertising or tracking SDKs.

Videos, Audio, and Captions

Videos you import, audio extracted for transcription, generated captions, project files, and exported videos are processed locally on your device for the core caption workflow.

We do not upload, inspect, store, or sell your source videos, project files, or exported videos. Audio leaves your device only if you turn on Cloud Recognition, described below. Exports remain in the app, Photos, Files locations you choose, or apps you share them with.

Optional Cloud Recognition

Cloud Recognition is an optional feature that transcribes speech in the cloud for faster, more accurate captions. Pro uses it within its monthly AI minutes; free accounts get up to 5 minutes per video and 60 minutes per month, after which captions are made on your device. It does not start until you agree in the in-app prompt, which names the providers, and you can turn it off at any time in Settings. When it is off, speech recognition runs entirely on your device.

When it is on, the app extracts the audio track of the video you are captioning, compresses it, and sends it with the spoken language (when known) and any glossary terms you entered through the Just Captions AI backend to a third-party speech recognition provider, which returns the transcript and word timings. Today that provider is ElevenLabs, with OpenAI or Groq used as a backup when ElevenLabs is unavailable. Video frames are never sent. Cloud Recognition is not offered in mainland China.

We do not keep the audio on our servers after the request completes.

Optional AI Correct and AI Translate

AI Correct and AI Translate are optional internet features. They do not start until you confirm the in-app network prompt — which names the third-party AI service used and the data sent — and tap the Start AI button.

When you use these features, the app sends caption text, nearby read-only caption context, glossary terms you entered, and the target language when translating to a third-party AI service (today: OpenAI for the non-China service path and DeepSeek for the China service path, with either provider used as a backup for the other when the primary one is unavailable; see "Third-Party AI Providers and Sharing" below) so it can return corrected or translated caption text. Your source video, video frames, and audio recordings stay on your device for these text-editing features.

Do not use AI Correct or AI Translate on confidential transcript text that you do not want processed by a third-party AI service or its model provider.

Face Data

Just Captions does not collect, detect, measure, analyze, identify, extract, or store face data, facial geometry, or biometric identifiers.

Videos that you choose to caption may incidentally contain a person's face, but the app does not use face detection or face recognition features to run its caption workflow. The app's core caption generation works from the video's audio and timing on your device.

Optional AI Correct and AI Translate do not send face imagery, video frames, or audio recordings to our AI service. They send caption text only, with limited text context needed to improve or translate subtitle wording.

Speech Processing

The app transcribes audio from videos you select so it can create captions. Core caption generation is intended to run on device and does not require an account.

Automatic captions, corrections, and translations may be inaccurate, so you should review caption text before saving or publishing a video.

Information We May Receive

Third-Party AI Providers and Sharing

For AI Correct and AI Translate, the text you choose to process is sent through the Just Captions AI backend to third-party AI providers. Today that provider path is OpenAI for the non-China service path and DeepSeek for the China service path. For service continuity, either provider may act as a backup for the other when the primary provider is unavailable, so text on the non-China path can reach DeepSeek and text on the China path can reach OpenAI. Each backup direction is used only when explicitly enabled.

For Cloud Recognition, audio is sent to ElevenLabs as described above, with OpenAI or Groq as a backup.

For the Emoji caption style, once you have agreed to AI network use, caption text is sent through the same backend to Groq, or to OpenAI when Groq is unavailable, which returns one emoji per caption. Without that agreement, emojis are picked on your device, using Apple's on-device model where available.

We do not sell your audio or transcript text. We share them only with the providers needed to return the transcription, correction, translation, or emoji result you requested.

Storage and Retention

We do not intentionally store your source videos, audio, exported videos, or transcript text on our own servers after AI processing completes. On our backend, audio and transcript text are handled in transit to complete the request.

We retain a short-lived anti-replay security nonce for about 6 minutes to protect the AI endpoint from abuse. Backend logs are intended to record operational events such as request counts, provider status, latency, and errors, not the transcript text itself.

On your device, your videos, captions, projects, and exports remain until you delete them or remove the app.

Purchases

Pro purchases, subscriptions, renewals, restores, cancellations, taxes, and refunds are handled by Apple through the App Store. We do not receive or store your payment card details.

Data Collection and Tracking

Just Captions does not track you across apps or websites. We do not sell personal data. App Store privacy answers should reflect optional AI transcript processing, support messages you choose to send, and Apple-provided diagnostics, purchase status, and analytics according to your settings.

Contact

For privacy questions, email abel0911@icloud.com.