# About SubLunio: AI subtitles in 110+ languages Your video, readable in more languages. Start with a free preview, keep the original video local, and continue only when the subtitles suit you. ## Start with your video and reach 110+ subtitle languages SubLunio is an AI subtitle generator for video files you already have. It helps turn dialogue into timed text, translate that text into 110+ target languages and export subtitle files you can keep. There are currently 118 target-language choices, from widely spoken languages such as English, Spanish, Chinese, Japanese, Arabic and Hindi to less commonly offered reading languages. A useful subtitle starts with the actual recording. A transcript copied from another cut can drift after an introduction, an advertisement or a different edit. SubLunio works from the audio in your selected file, giving you a starting point aligned to that version. You still control the source, chosen track, target language and final review. ## How online subtitle generation works The normal journey has three stages: choose the video and subtitle language, read a free preview, then continue with the complete subtitles if the result suits you. You can begin without registering. Select both the source language and the subtitle language before starting. Choose Auto for the source language if you want automatic detection. The browser prepares audio before uploading it. On the server, voice activity detection helps identify dialogue and segment work for the recognition provider. The application first transcribes and translates only the preview window. After verified payment, the same task continues with the remaining speech. This avoids asking you to upload the file again simply to move from a preview to the paid result, provided the temporary audio is still available. ## Local audio extraction without uploading the original video SubLunio reads the selected source file in bounded portions and prepares an MP3 in browser-managed temporary storage. The original video file and video frames are not part of the normal upload path. This is useful for large movies because video data is often much larger than the speech audio needed for subtitles. Local extraction should not be confused with offline transcription. The extracted audio is uploaded for server processing, and the speech and translation providers process the portions needed for their work. The initial workflow may upload the full extracted audio even though only the 5-minute preview is recognised and translated before payment. The Privacy Policy explains these separate stages and their retention rules. ## MP4, MKV, TS and videos larger than 2 GB People searching for an MP4 subtitle generator, an MKV subtitle tool or a way to generate subtitles from TS files often have the same problem: their original video is too large for a simple upload form. SubLunio's browser workflow does not impose a fixed 2 GB limit on the source video. It reads in portions rather than requiring the entire movie to be loaded into memory at once. Common intended inputs include MP4, MKV, TS, MOV and AVI. Successful processing still depends on the container, audio codec, file integrity and the browser's capabilities. Encrypted or DRM-protected content is unsupported. The current limit is 8 hours per account task or 3 hours as a guest, with up to 400 MB of extracted audio. Available device storage matters, so a large source file is not a promise of unlimited duration or compatibility with every recording. ## Anime subtitles with attention to names and context An Anime subtitle generator needs to handle more than isolated dictionary translations. Dialogue may contain character names, nicknames, recurring phrases and fictional place names. SubLunio's optional glossary lets you supply preferred forms, while the translation request includes neighbouring subtitle cues to help preserve the sense of the current line. This makes the tool useful when your own Anime episode has no suitable subtitle file, when you want a different reading language, or when you want to compare a translated SRT with the original-language text. Music, overlapping voices, stylised delivery and uncommon names can still cause errors. Preview representative dialogue and review important terminology; the output is AI-generated and is not an official or licensed subtitle release merely because it was generated here. ## For movies, series, lectures and creator videos A movie viewer may want translated dialogue without changing the video file. A learner may want the original line next to a translation. A researcher may need timed text for an interview they have permission to process. A teacher can prepare a lecture for a multilingual group, while a creator can add a separate caption track to a finished edit. These are different uses of the same portable subtitle workflow. The source matters. Music under speech, overlapping speakers, whispered dialogue and unfamiliar names can reduce recognition quality. A clean dialogue track and a glossary of preferred names help. If a video has several tracks, listen to the short sample first: subtitles for the director’s commentary will not describe the main dialogue. Choose one target language per task; additional language versions are separate full-video generations. ## Speech recognition and subtitle translation are different jobs SubLunio uses configured external AI services for automatic speech recognition. Recognition turns speech into timed source-language text. Translation is a separate step: the application sends subtitle lines, necessary context and glossary guidance through OpenRouter to the configured language model. The target language is therefore a distinct choice from the source-language detection setting. There are currently 118 selectable subtitle target-language options and 19 interface languages. Those counts describe the menus, not a guarantee of equal recognition coverage or equal translation quality in every language. A model's ability to understand a spoken language and its ability to write a target language are different constraints. Check your own preview, especially for dialects, code-switching and less common languages. ## A free 5-minute subtitle preview that starts with dialogue The preview covers up to 5 minutes starting at the first subtitle, rather than always starting at the first frame of the video. If the first subtitle is around 01:40, a full preview window extends up to 5 minutes from that point, unless the recording ends sooner. This makes the preview useful even when a movie begins with credits, music or a silent introduction. The preview page lets you read the translated cues, search the displayed text and show the original wording alongside it. Use it to check language, the selected audio track, names and the general reading quality before deciding to continue. It is a preview of generated subtitles, not a free download of someone else's subtitles or a promise that every later line will have identical quality. ## A price based on time, with no subscription Full generation costs USD $1 per video hour, rounded upward to the nearest cent and including tax. A 90-minute recording costs $1.50. The charge covers recognition and translation of the complete video, including the section you tried for free. You can choose a different output format for the same completed task without paying for another generation. After the preview, use existing account credit or pay the video’s remaining amount through Stripe. A card payment starts at $1.00; any excess remains in your balance for another recording. You can also add a whole-dollar amount from $1 to $100. Credit has no scheduled expiry. Direct full generation reserves an estimate and settles against measured audio duration before recognition begins. ## Portable SRT, VTT and ASS files SRT is a practical first choice for a local player or editing application: it stores subtitle timing and readable text with broad compatibility. VTT suits web video and platforms that accept caption tracks. ASS adds styling, which is useful when a bilingual result should show translated dialogue and a smaller original-language line. Browser previews are intended to help you judge the result before downloading. Choose translated, original or bilingual subtitles and the bilingual line order. When source and target match, the same line is not duplicated; export separate single-language files when viewers need to switch languages independently. After generation, export SRT, VTT or ASS subtitle files, or burn subtitles directly into your video in the desktop app. Burn-in creates a new video by default, with the option to replace the original where supported. Downloading a separate subtitle file does not change the original video. ## Start as a guest and keep the task when you sign in You do not need to create an account before trying the service. When you later sign in by email in the same browser, eligible guest tasks can be attached to the account. This is designed for the common journey of trying a subtitle preview first, then deciding that the result is worth keeping. It avoids making registration a prerequisite to discovering whether the tool works for your file. Account history retains subtitle tasks, not a permanent library of uploaded audio. Temporary audio normally has a 24-hour window, and the actual deadline appears in the task. Guest records normally have a 7-day retention period. Keep a private recovery key where provided, download results you wish to retain, and continue before the audio deadline if you want to avoid selecting the original file again. Server audio may be deleted earlier after full processing finishes, when you cancel, or when you delete the task. ## What progress indicators tell you Audio preparation and upload happen while your browser is involved, so keep the tab open for those stages. Once the server has accepted the upload and started the background task, you can leave the page and return to the task later. The interface distinguishes preparation, upload, speech detection, transcription and translation rather than treating them as one unexplained wait. The displayed percentage relates to a stage and is not an exact prediction of remaining time. Long videos, dense dialogue, model limits and network conditions can change the completion time. If something fails, the interface can offer a retry or recovery action appropriate to the available saved progress. A readable status should help you decide what to do next, not imply that every provider request will always succeed. ## A private workspace rather than a public media catalogue Your task belongs to your guest session or account. It is not published as a movie page, a shared subtitle library or a searchable recording. The service does not provide source videos. You are responsible for having the permissions needed to process the material, and for reviewing the rights relevant to sharing a resulting subtitle. Only the extracted audio is uploaded. Cloud transcription and translation still involve external processing, so local video handling does not mean the whole service is offline. The privacy policy describes providers, temporary server audio, local cache, account export and deletion. These details matter particularly when your recording contains private conversations or identifiable third parties. ## Batch subtitles on desktop: export files or burn subtitles into video The desktop application is designed for Windows 10/11 on x64 and macOS 12 or newer on Apple silicon. It connects to the same account and balance. Its persistent local queue prepares audio, resumes uploads and saves subtitle files beside the source or in a folder you choose. Selecting several target languages creates a separate chargeable task for each language. Desktop generation goes directly to the full result; try a representative video in the website preview first if you want to assess quality. Signed installers and release notes appear on the download page when a release is available. Until then, the browser workflow remains the available starting point. A task already processing in the cloud does not need the browser to remain open. After generation, export SRT, VTT or ASS subtitle files in batches, or burn subtitles directly into your videos. Before burn-in, check the result in the built-in player and adjust size, color, opacity and position. Supported video formats also offer selectable embedded subtitle tracks; their appearance depends on the subtitle format and player. Desktop checks each file before offering video export methods. Successful subtitle generation alone does not guarantee burn-in or embedded-track support. ## Begin with one representative scene Choose a video, set the language you want to read, and generate a preview. Use the local player to check the correct track, timing, names and line breaks. The free window starts at the first subtitle and runs continuously for up to 5 minutes, so an opening silence does not consume the useful beginning of the preview. If the result meets your needs, sign in and continue the same task. When full generation finishes, download the format that suits your player or platform and keep a local copy. The pricing page explains balance and settlement, the help pages explain playback and export, and the contact page provides a direct route to support when a file or payment needs attention. ## Keep exploring - [Pricing](https://sublunio.com/pricing) - [Help](https://sublunio.com/help) - [Privacy](https://sublunio.com/privacy) Source page: https://sublunio.com/about Last updated: 2026-10-08