Terms & Privacy

Your subtitles are private

Subtitles you generate are visible only to you, signed in on your account. We do not publish a browsable library; there is no public listing, no search page, and no sitemap that enumerates slugs.

Transcript cache

We compute an audio fingerprint of each upload and use it as a cache key. If the same audio has already been transcribed — by you or by another account — you receive the existing subtitle files instead of paying for a fresh translation. The match requires you to possess the source audio; we don't expose, list, or let anyone browse other users' uploads. The fingerprint is an opaque SHA-256 hash — not a filename, not audio, not the subtitle text.

What we store

• The filename you uploaded (including extension).
• The extracted audio, uploaded to Alibaba OSS via DashScope.
• The Japanese transcript (SRT) and English translation (SRT).
• File size, timestamps, and your display handle.
• An audio fingerprint (SHA-256) for dedupe.

We don't store the original video file on our servers — audio is extracted in your browser and uploaded directly to DashScope for transcription.

Processing providers

Audio and text are sent to third-party AI services to complete the transcription + translation pipeline:

• Alibaba Cloud DashScope (Qwen3-ASR) — Japanese transcription.
• DeepSeek — Japanese → English translation.
• Clerk — authentication.
• Upstash — Redis and QStash.
• Vercel — hosting.

Contact

Questions or takedown requests: notify@subs.rip.