# Dubnivo
> A mobile listening and learning companion that turns authorized foreign-language transcripts into translated narration, bilingual study material and private podcasts.

This is an optional public product summary. Linked pages and server-verified account quotas are authoritative. It is not a public developer API or a promise of search-engine visibility.

## Product and usage
- Users import text, TXT, SRT or VTT material they own or have permission to process, choose a supported source/target language and available voice tier, then generate translated speech.
- A YouTube link can retrieve an existing English transcript when available. Other source transcripts can be imported manually. Not every video has an accessible transcript.
- The result is a separate audio track with translated/original text. It does not replace the original video's soundtrack or provide lip synchronization.
- Generated audio is private to its account and kept online for 7 days. Replay and download within that retention window do not consume new characters. Changing the target language or voice requires a new job and corresponding quota.
- The browser player supports speed control, seeking, bilingual text and Markdown/VTT exports. Mobile background and lock-screen behavior depends on the device/browser.
- Long existing English captions can be retrieved up to 6 hours within text/response limits, then selected in parts of at most 20 source minutes and 11,500 characters. Generation still has a 12,000-input-character ceiling. Automatic channel generation produces only part 1, not an entire long video.
- Personal terminology and professional domains can guide translation; programming mode preserves fenced and inline code. Glossaries are account-owned and language-pair specific.
- Saved vocabulary retains the source text, translation and source link. Anki .apkg packages are generated locally, with up to 50 words and optional existing translated-sentence WAV clips; TSV/CSV exports remain available. English dictionary lookups send only the explicitly requested word to Free Dictionary API.
- AI questions use an owned, completed, still-retained transcript as context. Questions and replies are bounded by purchased/trial allowance, daily and service capacity, and the original shared generation/question provider cost budget. Attempt counts are upper limits, not a separately sold guarantee of all questions plus every character in all worst-case combinations. No image understanding or unrestricted chatbot is offered.
- Users may enable a revocable private RSS URL and explicitly export completed audio into MP3 without another text-to-speech call. Each source job has a lifetime maximum of three export attempts; already-ready MP3 playback is unaffected. Online files expire with the original seven-day retention; external podcast clients with the feed URL can access them until it expires or is revoked.
- Channel following uses the channel's public YouTube Atom feed, up to three channels/account. Explicit paid-account opt-in may generate the first part of newly published videos and optionally publish it to an enabled private feed. Polling is normally every six hours, subject to shared capacity; quotas, authorization and service availability still apply.
- The creator audio studio provides separate narration WAV and generated-audio-timed SRT/VTT exports. Voice cloning, speech recognition for videos without captions, video rendering, platform publishing, public API keys, SDKs and MCP access are not current product features.
- The website works without an extension. PWA share targets depend on browser support; iOS web access supports copy/open. Telegram and Safari helper source are prepared separately and require owner setup/distribution; they are not represented as connected production services.

## Pricing and billing
All amounts below are the product's listed USD prices, not underlying supplier costs. Taxes, when applicable, are shown at checkout.
Subscriptions are optional. One-time packs do not renew and expire after 365 days. Top-ups use fixed packs, not arbitrary recharge amounts. Billing is based on generated output characters and independent attempt/request limits, not video minutes.

| Plan | Price and payment period | Characters | Attempt/request limits |
| --- | --- | --- | --- |
| Standard monthly | $9.90 per month | 300,000 standard output characters per monthly release | 100 new generation attempts / 100 automatic transcript requests per month |
| Standard yearly | $99.00 per year, with 12 monthly quota releases | 300,000 standard output characters per monthly release | 100 new generation attempts / 100 automatic transcript requests per month |
| Premium monthly | $19.90 per month | 300,000 standard + 100,000 premium output characters per monthly release | 100 new generation attempts / 100 automatic transcript requests per month |
| Premium yearly | $199.00 per year, with 12 monthly quota releases | 300,000 standard + 100,000 premium output characters per monthly release | 100 new generation attempts / 100 automatic transcript requests per month |
| Standard pack | $4.90 one-time, valid 365 days after verified payment | 100,000 standard output characters per pack | 40 new generation attempts / 40 automatic transcript requests per pack |
| Premium pack | $5.90 one-time, valid 365 days after verified payment | 50,000 premium output characters per pack | 10 new generation attempts / 10 automatic transcript requests per pack |

- Yearly subscriptions are charged once per year and release quotas monthly for 12 cycles, rather than delivering the entire year's usage upfront.
- Subscription characters can carry into one further cycle. Generation attempts and automatic transcript request counts do not roll over. Canceling stops future renewals; already-paid rights keep their original expiration dates.
- A failed new generation releases reserved characters but still consumes a generation attempt. Paid automatic transcript requests are returned on failure; dispatched free-trial transcript attempts are not returned. Manual import and this account's transcript-cache hits do not consume another transcript request.
- New generation uses a quoted reservation before work and settles the actual output characters afterward. Public examples and local pricing estimates are not quota reservations or guaranteed final invoices.
- Quotas are granted only after server-side verification of Creem payment. Service and transcript capacity can limit new purchases or jobs.
- Original audio previews need no sign-in. Eligible verified Google accounts can request a one-time 7-day trial of 3,000 standard characters, 2 generation attempts and 1 English-transcript request, subject to availability and eligibility limits.

## Public pages
- [Home](https://getpodtube.com/en): Product and original audio previews.
- [Pricing and local pack calculator](https://getpodtube.com/en/pricing): All six listed prices, quotas, expiry rules and FAQ.
- [Pay-as-you-go packs](https://getpodtube.com/en/pricing/pay-as-you-go): One-time packs without a subscription.
- [How to use Dubnivo](https://getpodtube.com/en/docs): Transcript import, language/voice selection, playback and billing rules.
- [YouTube subtitles to audio](https://getpodtube.com/en/youtube-to-audio): Authorized existing transcripts for mobile listening.
- [中文使用指南](https://getpodtube.com/zh/docs)
- [繁體中文使用指南](https://getpodtube.com/zh-tw/docs)
- [Support](https://getpodtube.com/en/support): support@getpodtube.com.
- [Terms](https://getpodtube.com/en/terms)
- [Privacy](https://getpodtube.com/en/privacy)
- [Refund policy](https://getpodtube.com/en/refunds)
- [Sitemap](https://getpodtube.com/sitemap.xml)