Bilibili subtitles · TXT / SRT

Read and export Bilibili subtitles

Check for existing captions, then read by timestamp, search and export the sentences you need.

Guests have a daily reading allowance. Registered accounts can read existing subtitles without a reading limit; transcription and AI features use their own allowances.

Tip: replace youtube.com or bilibili.com in a video URL with vidleaf.app, then press Enter to open it in Vidleaf.

Drag this button to your bookmarks bar: Open in Vidleaf

From a link to a subtitle file

1

Paste a public video link

Enter a BV ID or video-page URL. For a multipart video, keep the p parameter so the correct episode opens.

2

Choose an existing caption track

Read platform captions directly when available. Decide whether to use speech transcription only after the check confirms there are no captions. A failed request does not mean captions are absent.

3

Select, copy or download

Select the sentences you need, copy their text or download selected TXT or SRT. SRT includes start and end times. Check the selection and display language before exporting.

Platform captions and speech transcription

Platform captions are existing manual or automatic subtitles. Speech transcription reads the audio and recognizes the words again, using transcription minutes. Both sources have timestamps and may contain mistakes or omissions.

Summaries, translation and speaker detection are separate choices. You do not need to enable them just to read captions.

What does an SRT excerpt look like?

These are the first three entries from an existing Chinese automatic caption track on Bilibili. The original text and times are preserved. This is neither a new Vidleaf transcription nor a download of the full track.

1
00:00:00,000 --> 00:00:01,800
我想我和伊利静都没想到

2
00:00:04,950 --> 00:00:08,530
这在一年前对我来说都是不可想象的

3
00:00:08,530 --> 00:00:09,670
我想对于他来说

English reading translation

  1. I think neither Yi Lijing nor I expected this.
  2. A year ago, this would have been unimaginable to me.
  3. I think, for that person…

This translation helps explain the excerpt. The third caption is an unfinished sentence. The name in the first Chinese caption appears to be misrecognized; check it against the video before quoting.

Source: Chen Luyu's first conversation with Yi Lijing — Slow Talk, episode 1, uploaded by 陈鲁豫-慢谈; Chinese platform-generated captions.

Possible recognition errors, including “伊利静”, remain in the source excerpt. Verify names and wording in the original video.

If a video cannot be read

SituationNext step
Missing, private or region-restricted videoCheck that it plays on Bilibili and confirm the BV ID and episode parameter.
The check confirms no platform captionsReview duration and available minutes in the transcription confirmation before proceeding.
Temporary platform failure, rate limit or network errorKeep the link and try again later. A fetch failure is not a reason to assume paid transcription is needed.

Check a real SRT export

Three captions selected in the saved Chinese reader, with selected and full SRT export controls
A saved Slow Talk video demonstrates exporting selected captions. This screenshot does not represent a new platform fetch or an accuracy test.

Follow the illustrated steps from pasting a link to downloading. For a first attempt, export three captions and inspect their words and timestamps in a text editor before exporting more.

Choose TXT, SRT or quotation Markdown

Your purposeFormatCheck before exporting
Edit the text in a notes appTXTChoose original, translated or bilingual text and confirm the selected sentences.
Use a subtitle file in a playerSRTUse the same video version and episode, then check synchronization at several timestamps.
Keep a quotation and its source for writingSelected Markdown or copy quotationCheck timestamps, attribution and machine translations. A summary is not a direct quotation.

For a few sentences, select them and use a selected export. For the full track, use full TXT or full SRT. After changing pages, check the selected count so earlier selections are not included accidentally. SRT preserves caption timing; exporting does not turn fragmented or overlapping source captions into complete sentences.

Find the words and check their context

Search for a name or a distinctive word, then read the surrounding captions. Simplified Chinese queries can match Traditional Chinese captions, but a misrecognized name may still be missing. Try a nearby word or locate the section by chapter and timestamp. Open the original video at that time and confirm the context.

Automatic captions often split a sentence across several rows. Include qualifications such as “if”, “may” and “not”; select neighboring captions when needed. Check names, numbers, units and what a statement refers to. Machine translations aid understanding but do not replace the source. When a translation is missing, read the original or explicitly request a translation instead of filling in words and presenting them as a quotation.

The reader offers language and Chinese-script display choices. Check the current mode before exporting. A source label explains where captions came from; it is not a guarantee of human-verified accuracy. Verify even manually authored captions before quoting them publicly.

Common questions

How do I extract Bilibili captions?

Use Vidleaf in your browser; no extension is required. Paste a BV ID or video link, check for existing captions, choose a track, search or select sentences, then export TXT, SRT or quotation Markdown. Reading an existing track does not require speech transcription.

Can I use it without installing an extension?

Yes. Open this page in a browser and paste a public video link or BV ID. When captions are available, you can read, search, select and export them. If the check confirms there are no captions, review transcription minutes and the audio-processing information.

How can I transcribe a video without captions?

First distinguish confirmed absence of captions from a failed read. After captions are confirmed absent, sign in, review the duration, available minutes and audio-processing information, and confirm transcription. Retry or check access first when a fetch is rate-limited or fails. Verify recognized names and sentence boundaries against the audio.

Can I export SRT?

Yes. Select captions and choose the selected SRT download, or use full SRT for the current track. Start and end times are retained; recognition errors and sentence breaks are not automatically corrected. Check timing and source wording before using the file.

What are the steps?
  1. Paste a BV ID or video-page URL; retain p for multipart videos.
  2. Check existing captions and confirm the title, episode and language track.
  3. Search or select sentences, then choose a selected or full export.
  4. Download TXT, SRT or quotation Markdown and check the content and source.
Can I read captions without signing in?

Yes. Read, search and export subtitled videos without signing in, with a daily limit on new reads. Signed-in users can read subtitled videos without a limit, use transcription, summaries and translation within their allowances, and save history and personal labels. Email code or Google; no phone number needed.

Reading existing captions and recognizing audio are separate operations. Start with the existing track when you only need TXT or SRT.

What if no captions are available?

After the check confirms this, sign in and review duration, available minutes and audio-processing information before confirming transcription. A failed fetch or restricted video does not prove there are no captions. Verify transcription against the audio. See allowances and plans.

Are captions and my reading history public?

Personal tasks, history and speaker labels are not published as public articles. Video captions, translations and summaries are cached per video so other readers of the same video may reuse them, including guests when saved results are available. They are not private personal notes.

Ordinary reading results do not get their own search-indexed article. See the privacy policy for handling and retention, and retain attribution when quoting.

Continue reading