# VideoScript — Video & Podcast Transcripts for Notes and AI URL: https://videoscript.uk/ Make any public episode usable. Transcripts from public videos & podcasts — ready for AI & notes. Public video or podcast URLSign up and continue120 free Credits · No card required CreatePaste one public link. ProcessWe extract and transcribe. KeepCopy or download the complete transcript. Supported platforms YouTube Bilibili Apple Podcasts Xiaoyuzhou Spotify The transcript stays yours to use. Read it here, copy it into your notes or AI, or export Markdown, SRT, VTT, and TXT. Credits are minutes, not a subscription. Try with 120 free Credits Frequently asked questions - What links can I use? A specific public YouTube video, Apple Podcasts episode, Spotify episode (via the publisher's public RSS feed), Bilibili video, or Xiaoyuzhou episode. Private, paid, DRM-protected, login-only, and region-restricted content is not supported. - Is this a summary? No. VideoScript gives you the full timecoded transcript. You can copy it or export it as TXT, Markdown, SRT, or VTT, then summarize it with the AI or notes tool you already use. - How much does it cost? One credit covers one minute of media. New accounts get 120 free credits. One-time Credit packs and 30-Day Passes start at $9.99, and the largest 30-Day Pass works out to about $0.55 per hour. Packs never expire; each Pass is valid for 30 days and never auto-renews. All prices and credit amounts are listed on the pricing page — the single source of truth for what you pay today. - Do 30-Day Passes renew automatically? No. A 30-Day Pass is a one-time purchase. Nothing bills you again unless you buy another Pass yourself. - How long does a transcript take? Transcription runs at about 50x real time, so one hour of audio is typically transcribed in under two minutes once processing begins. Total turnaround also depends on the source and queue. - What happens to the audio? Audio is used only while your transcript is being processed and lives on temporary storage. VideoScript does not keep it or offer it as a download. # Pricing — VideoScript URL: https://videoscript.uk/pricing/ Simple pricing for every transcript. One Credit covers one minute of actual media. Choose permanent hours or a lower-rate 30-Day Pass. Have a community code?Enter it to check out at your community's price.Enter your code▾ How to buy 30-Day Pass More hours for a focused month. Priority processing and AI summaries are coming. One payment · Valid for 30 days · Never auto-renews One-time packs Hours that wait until you need them. One payment · Never expires Loading plans…Plans appear once the page finishes loading. How pricing works - One Credit covers one minute of media, rounded up to the next whole minute. - New accounts start with 120 free Credits — no card required. - One-time packs never expire; the hours wait until you need them. - A 30-Day Pass is one payment, valid for 30 days, and never auto-renews. - Failed transcripts do not use Credits. How VideoScript pricing works, in detail → # Terms of Service — VideoScript URL: https://videoscript.uk/terms/ Terms of Service Last updated: 2026-09-09 Welcome to VideoScript. These Terms of Service ("Terms") govern your use of the VideoScript website and service (the "Service"), operated by VideoScript ("we", "us"). 1. What VideoScript does VideoScript turns a public video or podcast episode into a complete, timecoded transcript. You paste a link; we extract the audio or captions and transcribe it. You can read, copy, and export the transcript as TXT, Markdown, SRT, or VTT. Supported sources are public YouTube videos, public Apple Podcasts episodes, public Xiaoyuzhou episodes, public Bilibili videos, and public Spotify podcast episodes (via the publisher's public RSS feed). 2. Your account You must sign in to submit a task — with a Google account or an email verification code. You are responsible for keeping your account secure and for everything that happens under it. 3. Acceptable use You may only submit content that is publicly available and that you are allowed to use. You must not use the Service for content that is private, paid, DRM-protected, or behind a login, and you must not attempt to bypass any access control, region, or platform restriction. You must not scrape, probe, or abuse the Service, interfere with its operation, or resell access. We may suspend or terminate accounts that violate these rules. 4. Credits Credits are minutes: 1 Credit covers 1 minute of actual media. A new account receives a one-time 120 Credit trial. When you submit a task we reserve the estimated Credits, then settle the final amount from the actual media length (rounded up to the nearest minute, minimum 1). A task that does not complete releases its reservation — you are only charged for successful transcripts. 5. Purchases & 30-Day Passes You may buy Credit packs (one-time, never expire) and one-time 30-Day Passes. Each Pass grants its stated capacity for 30 days; unused Pass capacity expires with that Pass and does not carry over. Passes do not renew automatically, and you are charged only the amount you approve at checkout. All payments are processed by Stripe; your card details go to Stripe and never reach us. Refunds are described in the Refund Policy. 6. Your content You keep the rights to the links you submit and the transcripts we produce for you. We store your tasks and transcripts so you can return to them; you can delete them at any time. Temporary audio is used only to produce the transcript and is not kept or offered as a download. 7. Third-party content Transcripts of third-party content are provided as a convenience for your personal use. You are responsible for how you use a transcript and for complying with the source platform's terms and the rights of the content owner. 8. Availability & accuracy The Service is provided "as is" and "as available". We work to keep it reliable, but we do not guarantee uninterrupted availability, and transcripts are machine-generated and may contain errors. We are not responsible for what a third-party platform does to a link. 9. Limitation of liability To the maximum extent permitted by law, VideoScript is not liable for indirect, incidental, or consequential damages, or for any loss of data or profit, arising from your use of the Service. Our total liability is limited to the amount you paid us in the 12 months before the claim. 10. Termination You may stop using the Service and delete your account at any time. We may suspend or terminate access for a violation of these Terms or to protect the Service. 11. Changes We may update these Terms. Material changes will be posted on this page with a new date. Continuing to use the Service after a change means you accept it. 12. Governing law These Terms are governed by the laws of Hong Kong, without regard to conflict-of-law rules. Any dispute will be resolved in the courts of Hong Kong. 13. Contact Questions about these Terms: support@videoscript.uk. # Privacy Policy — VideoScript URL: https://videoscript.uk/privacy/ Privacy Policy Last updated: 2026-09-09 VideoScript ("we", "us") operates the transcript service at videoscript.uk. This policy explains what we collect, why, and your choices. 1. Data we collect - Account data: your email address (or your Google account) and any name you provide, handled by our sign-in provider Clerk. - Content you submit: the public link you paste. - Generated content: the transcript and metadata we produce (title, duration, source, language, timestamps). - Usage data: technical logs (IP address, browser, pages visited, session id) needed to run and secure the Service. - Billing data: payment is processed by Stripe. We receive a payment reference and amounts; we never see or store your card details. 2. How we use it - To create and return your transcript, and to keep your history and Credit ledger. - To provide support and respond to you. - To keep the Service secure and prevent abuse. 3. Who we share it with We use trusted processors: Clerk (authentication), Stripe (payments), Cloudflare (hosting and delivery), and transcription providers. We share only what each needs to do its job. We do not sell your personal data. 4. Retention - Your account, tasks, and transcripts are kept while your account is active; you can delete a task and its transcript at any time. - Temporary audio is deleted after each task is processed and is never kept as an asset. - Payment records are kept as long as required for accounting and legal purposes. 5. Your rights You may access, correct, export, or delete your data by using the Service or contacting us. Deleting a task removes its transcript. 6. Children The Service is not intended for children under 13, and we do not knowingly collect their data. 7. International transfers We and our processors operate in multiple countries. Where required, transfers rely on appropriate safeguards. 8. Cookies & local storage We use essential storage (for example, to remember your session and a pending invite or community offer code). Sign-in uses Clerk's cookies. We do not use advertising trackers. 9. Changes We will post any changes here with a new date. 10. Contact support@videoscript.uk. # Refund Policy — VideoScript URL: https://videoscript.uk/refunds/ Refund Policy Last updated: 2026-09-09 Current status Purchases are open. One-time Credit packs and one-time 30-Day Passes are sold through Stripe, and the rules below apply from the day of purchase. Nothing on this site bills you automatically — a 30-Day Pass never renews itself, so another 30 days only begin if you buy another Pass. Trial Credits are free and cannot be exchanged for cash. Credit packs - No charge for failures. You are only charged for successful transcripts. A task that fails releases its reservation and costs nothing. - Spent Credits are not refunded. Once a transcript has been delivered for a Credit, that Credit is used and not refundable. - Unused Credits. If you bought a pack and have not used it, you may request a refund within 14 days of purchase by emailing support@videoscript.uk. Packs never expire, but refunds apply to unspent Credits only. 30-Day Passes - No charge for failures. You are only charged for successful transcripts; a failed task releases its reservation and costs nothing. - Used quota is not refunded. Once a transcript has been delivered using your 30-Day Pass, that part is used and not refundable. - Unused Pass. If you bought a 30-Day Pass and have not used it, you may request a full refund within 14 days of purchase. - No auto-renewal. 30-Day Passes are one-time purchases and never renew automatically — you are only ever charged the amount you approved at checkout. Shared - Payment errors. If you are charged twice or the wrong amount, contact us and we will correct it. - Chargebacks & abuse. Refunds are processed back to the original payment method. We may decline refunds where there is evidence of abuse or a violation of the Terms. - Statutory rights. Nothing in this policy limits any refund rights you have under applicable consumer law. Contact support@videoscript.uk. # How it works — VideoScript URL: https://videoscript.uk/how-it-works/ How it works By Erwin Wu · Updated 2026-09-21 VideoScript turns public video URLs, podcast episodes, or your uploaded audio files into complete, timecoded transcripts. Paste a link or upload a file, and keep a searchable transcript you can read, copy, or export into your notes and AI tools. 1. Submit your media: link or audio file You have two easy ways to get started: - Paste a public link — Copy the URL of any public YouTube video, Apple Podcasts episode, Spotify episode, Bilibili video, or Xiaoyuzhou episode and paste it into the workspace. - Upload an audio file — Upload voice memos, recorded lectures, or podcast tracks directly. Supported formats include MP3, M4A, AAC, WAV, OGG, OPUS, and FLAC (up to 100 MB and 120 minutes per file). Files are securely staged in private storage and physically deleted upon job completion. New accounts receive 120 free Credits on signup — covering 2 full hours of media with no credit card required. 2. Fast, high-accuracy transcription VideoScript extracts the audio, identifies speech segments, and aligns every line with an exact timestamp ([MM:SS]). Processing runs at roughly 50x real-time speed — a 60-minute discussion is ready in about 1.5 to 2 minutes. 3. Export and use your transcript Read the transcript directly in your browser, copy plain text or Markdown, or download standard subtitle files (SRT / VTT). Failed tasks never consume Credits. Ready to try VideoScript? Start with 120 free Credits and transcribe your first video, podcast, or audio file right now. Open Workspace → Credits and pricing - One Credit = one minute of media. New accounts get 120 free Credits once. - One-time packs: permanent hours that never expire, waiting until you need them. - 30-Day Pass: lower per-hour rate, valid for 30 days, never auto-renews. See the full breakdown on the pricing page or read our pricing guide. For detailed platform walkthroughs, see our guides on YouTube transcripts, podcast transcripts, or feeding transcripts into AI. # Supported sources — VideoScript URL: https://videoscript.uk/sources/ Supported sources By Erwin Wu · Updated 2026-09-21 VideoScript extracts and transcribes speech from five major public media platforms as well as direct audio file uploads. Paste a link or upload an audio file to receive a complete, timecoded transcript in minutes. Supported public platforms - YouTube — Public videos, lectures, and interviews. If official captions are present, VideoScript uses them instantly; if captions are missing or disabled, VideoScript automatically transcribes the audio track. See how to get a YouTube transcript. - Apple Podcasts — Public podcast episodes. VideoScript extracts the full audio track, bypassing the 200-word limit imposed by the native Apple Podcasts reader. See how to transcribe podcasts. - Spotify — Public podcast episodes, resolved through the publisher's public RSS feed. Delivers downloadable Markdown, TXT, and SRT files not available in Spotify's app. See transcripts for notes and AI. - Bilibili — Public single videos. VideoScript processes the audio stream to generate complete Chinese transcripts suitable for translation or research. - Xiaoyuzhou — Public podcast episodes across Chinese podcasting channels. Direct local audio uploads Have a personal recording, lecture capture, or voice memo? You can upload audio files directly into VideoScript: - Supported formats: MP3, M4A, AAC, WAV, OGG, OPUS, and FLAC. - File constraints: Files up to 100 MB in size and up to 120 minutes in duration. - Privacy & retention: Audio files are uploaded to private temporary cloud storage, processed immediately, and physically deleted upon task completion. VideoScript does not retain your audio. What is not supported VideoScript strictly processes public content and user-authorized audio files. Private videos, subscriber-only paywalled podcasts, DRM-protected streams, and content requiring login or geoblocking bypass are not supported. No user passwords, session cookies, or proxies are ever used. Have a link or recording ready? Turn any supported episode or audio file into a clean, searchable transcript. Get 120 free minutes upon signup. Start Transcribing Free → What you get with every transcript Every transcript is complete (never truncated into a brief summary), carries exact per-line timestamps ([MM:SS]), and can be exported as Markdown, TXT, SRT, or VTT. For step-by-step guides on using transcripts, see YouTube transcripts, podcast transcripts, or feeding transcripts into Claude, ChatGPT & NotebookLM. # Spotify Podcast Transcript Generator — Export TXT, Markdown & SRT | VideoScript URL: https://videoscript.uk/spotify-podcast-transcript/ Export Spotify Podcast Transcripts: Download TXT, Markdown & SRT By Erwin Wu · Updated 2026-09-21 Paste any public Spotify podcast episode URL to extract a complete, timecoded transcript: Public Spotify podcast episode URLGet Full Transcript →120 free Credits on signup (120 minutes of audio) · No credit card required · ~50x real-time transcription Why Spotify doesn't let you download podcast transcripts Spotify includes an automated read-along transcript feature for select podcast episodes. While useful for following along in real time, the interface is deliberately restrictive: - No export button — You cannot download, copy in bulk, or save the transcript as a file. - Inconsistent availability — Many shows lack automated transcripts or have transcripts disabled by the creator. - No structured formats — There is no option to export formatted Markdown for personal notes (Obsidian/Notion) or standard subtitle formats (SRT / VTT) for editing. How VideoScript extracts complete transcripts from Spotify VideoScript bypasses these in-app limits through an open, transparent pipeline: - Public RSS resolution — When you paste a Spotify episode URL, VideoScript matches the episode against the publisher's public RSS syndication feed. - Direct audio extraction — It downloads the full audio recording directly from the creator's host server at original quality. - High-fidelity speech recognition — Our automated transcription pipeline identifies spoken phrases and attaches precise [MM:SS] timestamps to every sentence. Because this process operates through the creator's open RSS feed, you never need a Spotify premium account or private login credentials. What you can export - Markdown (.md) — Structured paragraphs with compact timestamp markers, ideal for Obsidian, Notion, and Logseq. - Plain Text (.txt) — Clean sentences without code markers for rapid proofreading, blogging, and drafting. - Subtitles (.srt / .vtt) — Industry-standard subtitle tracks with microsecond timing for Premiere Pro, Final Cut, and DaVinci Resolve. Supported vs. unsupported Spotify content - Supported — Public podcast episodes distributed through standard podcast RSS feeds (representing over 95% of podcasts listed on Spotify). - Unsupported — Subscriber-only paid shows, Spotify-exclusive DRM-protected audiobooks, music tracks, and private patron feeds. VideoScript never circumvents paywalls, DRM, or access controls. Step-by-step: How to get your transcript - Copy the link: In the Spotify app or web player, navigate to the episode, click the three dots (...) or share icon, and select Share → Copy link to episode. - Paste into VideoScript: Enter the link in the input box above or in our workspace. - Extract & transcribe: Processing runs at approximately 50x real-time speed. A 60-minute interview is transcribed in under two minutes. - Export your file: Choose Markdown, TXT, SRT, or VTT. Ready to transcribe a Spotify podcast? Start with 120 free Credits and get a complete, exportable transcript in minutes. No credit card needed. Transcribe Spotify Episode Now → Downstream workflows: Feeding Spotify episodes into AI Exporting your podcast into Markdown allows you to feed deep, two-hour conversations directly into Claude (with its 200k token window), ChatGPT, or Google NotebookLM. Learn more in our guide on feeding transcripts into Claude, ChatGPT & NotebookLM. For video sources, see our YouTube transcript guide. Frequently asked questions - Do I need a Spotify account to transcribe an episode? No. VideoScript resolves the public syndication feed provided by the show's creator, so you don't need to be logged into Spotify. - Can I download the transcript as a PDF? You can export your transcript as a .txt or .md file, which can be opened and printed or saved as a PDF in any text editor or browser. - How much does it cost? Every new account starts with 120 free Credits (1 Credit = 1 minute of audio). Paid packages start at $9.99 for 18 hours of one-time credits that never expire. See the pricing page for full details. - What happens if a podcast episode cannot be resolved? If an episode is private, geoblocked, or DRM-restricted, VideoScript detects this cleanly before transcription begins. Failed tasks never consume your Credits. # Bilibili Video Transcript Generator — Extract Chinese Speech to Text | VideoScript URL: https://videoscript.uk/bilibili-transcript/ Bilibili Video Transcript Generator — Extract Chinese Speech to Text By Erwin Wu · Updated 2026-09-21 Paste any public Bilibili video URL (BV ID) to generate a complete, timestamped transcript: Public Bilibili video URLGet Bilibili Transcript →120 free Credits on signup (120 minutes of media) · High-accuracy Chinese speech-to-text · No credit card required Why extracting transcripts from Bilibili is challenging Bilibili is one of the world's richest video archives for Chinese lectures, tech walkthroughs, academic symposiums, and cultural documentaries. However, accessing transcripts as a student or researcher is difficult: - Captions are rarely downloadable — Even when creators provide subtitles, Bilibili provides no export button for viewers to save text or subtitle files. - Many videos lack subtitles — A substantial proportion of technical tutorials and university talks on Bilibili have no CC subtitle tracks. - Language barriers — Non-native Chinese speakers and international researchers struggle to follow fast-paced speech without structured text to feed into translation tools. Direct audio transcription without relying on B站 captions VideoScript approaches Bilibili differently from simple caption scrapers: - Direct audio processing — Rather than failing when Bilibili captions are missing, VideoScript streams the public audio track directly and runs speech-to-text transcription. - Accurate Chinese recognition — Our pipeline handles Mandarin dialects, technical nomenclature, and mixed Chinese-English technical terms with high precision. - Exact timestamping — Every sentence is tagged with [MM:SS], making citation and reference verification seamless. The research workflow: Translating Bilibili content with Claude & DeepL Once VideoScript extracts your clean Markdown transcript: - Export the transcript as Markdown (.md). - Upload or paste the text into Claude 3.7 or DeepL. - Prompt the AI to translate paragraph by paragraph while preserving the [MM:SS] timestamps. This gives you an authoritative bilingual research document that you can cross-reference back to the original video at any second. What you can export - Markdown (.md) — Clean Markdown formatted for Obsidian, Notion, and translation prompts. - Plain Text (.txt) — Unformatted text for rapid reading or archiving. - Subtitles (.srt / .vtt) — Standard subtitle files for video editing or overlay. Supported vs. unsupported Bilibili links - Supported — Public single videos (e.g., https://www.bilibili.com/video/BV...). - Unsupported — Multi-part pagelists (?p=2), anime/bangumi requiring VIP subscriptions, live streams, and login-protected content. VideoScript respects platform boundaries and never bypasses paywalls or regional blocks. Transcribe a Bilibili video right now Paste your Bilibili link and get a complete Chinese transcript in minutes. Start with 120 free Credits — no card required. Transcribe Bilibili Video Now → Frequently asked questions - Do I need a Bilibili VIP account? No. VideoScript transcribes public Bilibili videos directly without requiring any login credentials. - What if the Bilibili video has no CC subtitles? VideoScript does not rely on Bilibili CC subtitles; it transcribes the spoken audio directly using high-fidelity speech-to-text models. - How fast does it transcribe? Audio processes at roughly 50x real-time speed. A 40-minute Bilibili documentary or lecture is typically ready in under a minute. # Guides — VideoScript URL: https://videoscript.uk/guides/ Guides By Erwin Wu · Updated 2026-09-14 Short, practical guides on getting transcripts and putting them to work. Each one takes about a minute to read and links to the pages it mentions, so you can follow every step without leaving the site. All guides - YouTube transcriptsWhether you are quoting a specific timestamp, scanning a 2-hour conference talk, or preparing source text for an AI prompt, accessing a complete YouTube transcript should take seconds. Here is how to view native transcripts on YouTube, how to copy clean text without clutter, and how to extract transcripts when captions are missing or disabled. - Podcast transcriptsPodcast episodes hold valuable research, interviews, and insights, but audio is inherently unsearchable. Whether you need to cite a timestamped quote, study a lecture, or feed a 90-minute conversation into Claude or Obsidian, a complete transcript makes audio usable. Here is how to transcribe podcast episodes from Apple Podcasts, Spotify, and local audio files — and how to overcome platform export limits. - AI-ready transcriptsLarge language models like Claude, ChatGPT, and NotebookLM are unmatched at summarizing arguments and extracting research takeaways. However, they lack ears: paste a YouTube link or a podcast episode into a chat window, and the model will decline. Here is how to bridge the gap — turning any public video or podcast into clean, token-efficient Markdown, and using it as grounded context in your favorite AI. - Transcript vs. summary: which do you need?A transcript is every word, in order, with timestamps. A summary is a shorter retelling someone chose for you. If you need to quote someone exactly, search what was said, or build notes you can trust, start with the transcript — you can always summarize from it, but you cannot recover the exact wording from a summary. - Turn a lecture into searchable notesTaking notes from a 90-minute university lecture or technical talk is frustrating when you have to pause every thirty seconds to scribble down equations or arguments. A complete, timestamped transcript converts hours of spoken video and audio into structured text you can search, highlight, and link directly into your second brain. Here is how to turn YouTube videos, recorded lectures, and podcasts into clean Markdown notes in Obsidian and Notion. - How VideoScript pricing worksVideoScript bills by the minute, not by subscription: one Credit covers one minute of media, rounded up to the next whole minute, and Credits only leave your balance when a transcript succeeds. You can buy hours once and keep them, or take a 30-Day Pass for a lower per-hour rate. Here is how both work and how to tell which one fits. - How many tokens in a 1-hour podcast?When feeding long podcasts, interviews, and lectures into large language models, token consumption dictates context fit, processing speed, and API cost. But how many tokens does spoken audio actually produce per minute or per hour? Here is a rigorous benchmark measuring speech rate, token density, and formatting overhead across modern frontier models including GPT-5.6, Claude Opus 4.6 / 4.8, and Gemini 3.7 / 3.8. New here? Start with how VideoScript works or check the supported sources. # About — VideoScript URL: https://videoscript.uk/about/ About VideoScript Updated 2026-09-14 VideoScript is built and run by two people: Erwin Wu, who writes the product, and Kimberly, who looks after marketing and user growth. This page names them, says what each one does, and shows where to find them. The team Erwin Wu designs and writes the product: the extractors that pull audio out of public links, the transcription pipeline behind them, the credits and checkout logic, and the pages you are reading. You can find him on X: @ErwinWu000 and GitHub: DrErwin. Kimberly handles marketing and user growth for VideoScript. You can find her on X: @king1818888. Why the product exists Talks, interviews and lectures are easy to watch and hard to use — you cannot search a video, quote a podcast, or paste an hour of audio into your notes. VideoScript turns one public link into a complete, timecoded transcript you can read, search, quote and export, so the words outlive the recording. Contact support@videoscript.uk New here? Start with how VideoScript works, read the guides, or check pricing. # How to Get a YouTube Video Transcript (Even Without Captions) — VideoScript URL: https://videoscript.uk/guides/youtube-transcript/ YouTube transcripts By Erwin Wu · Updated 2026-09-21 Whether you are quoting a specific timestamp, scanning a 2-hour conference talk, or preparing source text for an AI prompt, accessing a complete YouTube transcript should take seconds. Here is how to view native transcripts on YouTube, how to copy clean text without clutter, and how to extract transcripts when captions are missing or disabled. How to view native transcripts on YouTube If a video already has captions enabled, YouTube provides a built-in transcript reader: - Desktop: Below the video player, click ...more in the description box, scroll to the bottom, and click Show transcript. - Mobile: Expand the video description box, scroll past the chapters, and tap Show transcript. - The transcript panel will open, highlighting phrases in real time as the video plays. How to copy YouTube transcripts cleanly Highlighting and copying text from YouTube's native panel often results in broken sentences and unwanted timestamps cluttering your notes. - Hide timestamps: Click the three vertical dots (⋮) in the top-right of the transcript panel and select Toggle timestamps. - Limitation: Even without timestamps, copied text lacks proper paragraph structure, punctuation may be missing, and there is no one-click export to Markdown or subtitle formats. What to do when captions are missing or disabled If you encounter a video where the Show transcript button is greyed out or missing entirely, the creator may have disabled automatic captions, or background music may have blocked YouTube's automated system. This is where VideoScript steps in with a captions-first fallback architecture: - When a public YouTube URL is submitted, VideoScript first checks for existing captions for maximum speed. - If captions are disabled or unavailable, it automatically switches to high-fidelity audio transcription — extracting speech directly from the audio stream. - You get a complete, timecoded transcript without needing to download large video files or run manual converters. Export formats: Markdown, TXT, SRT, and VTT Rather than manually cleaning copied text, VideoScript lets you export directly to your preferred workflow: - Markdown (.md) — Formatted with clear time markers for Obsidian, Notion, and AI models. - Plain Text (.txt) — Clean text without formatting for quick reading or drafting. - Subtitles (.srt / .vtt) — Standard subtitle files for Premiere, DaVinci Resolve, or video clips. Next steps: Feeding your transcript into AI Once you have exported your clean transcript, you can turn it into executive summaries, research notes, or study guides. Read our guide on feeding transcripts into Claude, ChatGPT, and NotebookLM. Related guides - Podcast transcripts - AI-ready transcripts - Transcript vs. summary: which do you need? - Turn a lecture into searchable notes - How VideoScript pricing works - How many tokens in a 1-hour podcast? How VideoScript works · Pricing # How to Transcribe a Podcast Episode (Apple Podcasts, Spotify & Audio) — VideoScript URL: https://videoscript.uk/guides/podcast-transcript/ Podcast transcripts By Erwin Wu · Updated 2026-09-21 Podcast episodes hold valuable research, interviews, and insights, but audio is inherently unsearchable. Whether you need to cite a timestamped quote, study a lecture, or feed a 90-minute conversation into Claude or Obsidian, a complete transcript makes audio usable. Here is how to transcribe podcast episodes from Apple Podcasts, Spotify, and local audio files — and how to overcome platform export limits. The podcast transcription gap: why listening isn't enough Long-form podcast interviews and educational series contain dense knowledge, but audio is fundamentally linear. Scrubbing back and forth across a 75-minute episode to locate a single quote, statistic, or argument wastes precious time. Having a complete, timestamped transcript unlocks three critical advantages: - Instant searchability — Ctrl+F to find specific concepts or key topics instantly. - Verifiable citation — Every line is marked with [MM:SS] timestamps so any excerpt remains checkable against the original recording. - Downstream integration — Export clean text or Markdown directly into Obsidian, Notion, or frontier AI models like Claude and ChatGPT. Apple Podcasts transcripts: native read-along vs. export limits Starting in iOS 17.4, Apple introduced native read-along transcripts in the Apple Podcasts app across iPhone, iPad, and Mac. While convenient for casual listening, Apple's implementation imposes severe friction for research and productivity: - The 200-word copy limit — Apple explicitly restricts text selection: you can only copy up to 200 words at a time, and the copied snippet forces an attribution link back to the show. Exporting a 15,000-word episode requires copying dozens of times by hand. - No Windows or web access — Apple does not provide a web player with transcript export or a Windows client for reading transcripts. - Episodes missing transcripts — Apple's official documentation notes that episodes over 10 hours are not transcribed, newly released episodes may take over 24 hours to generate, and episodes with background music, poor audio quality, or crosstalk often have transcripts withheld entirely. How VideoScript fixes this: Paste any public Apple Podcasts episode URL into VideoScript. VideoScript extracts the full audio track directly from the publisher and generates a complete, uninhibited transcript for the entire episode — no 200-word barriers, and full export to Markdown, TXT, SRT, or VTT. Spotify podcast transcripts: app-only reader vs. file export Spotify provides real-time scrolling transcripts for select podcast episodes within its mobile and desktop apps. However, Spotify treats transcripts as an in-app companion rather than an exportable document: - Listeners cannot download, copy, or export the transcript as a text file. - Availability depends entirely on whether the creator enabled the feature or if automated transcription processed the episode. - There is no way to export standard subtitle files (SRT / VTT) or Markdown files for your personal knowledge base. How VideoScript handles Spotify: VideoScript resolves the episode by checking the creator's public RSS feed. It pulls the audio cleanly without touching Spotify's private login systems and produces a timecoded transcript you can download immediately. (Note: Spotify-exclusive DRM-protected content or subscriber-only paywalled episodes cannot be extracted.) Transcribing local audio files & voice recordings What if your podcast or interview isn't published on a streaming platform yet? If you have recorded a meeting, voice memo, lecture, or rough interview track, you can upload the file directly: - Accepted formats: MP3, M4A, AAC, WAV, OGG, OPUS, and FLAC. - Size & duration: Up to 100 MB per file and up to 120 minutes in length. - Physical deletion: Your audio is held in secure private staging during processing and physically purged the moment transcription concludes. Step-by-step: transcribing your episode with VideoScript - Step 1: Get your URL or file — Copy the public share link from Apple Podcasts, Spotify, or Xiaoyuzhou, or prepare your audio recording. - Step 2: Submit to VideoScript — Paste the link or drag and drop your file into the VideoScript workspace. New accounts receive 120 free Credits on signup — enough for two full hours of audio without entering a credit card. - Step 3: Process at 50x speed — Transcription runs at roughly 50x real-time speed. A 60-minute podcast episode finishes in approximately 90 to 120 seconds. - Step 4: Export your preferred format — Select Markdown (.md) for Obsidian and Notion, Plain Text (.txt) for quick reading, or Subtitles (.srt / .vtt) for video editing. Transcribe your podcast episode Get a complete, timecoded transcript without the 200-word copy limit. New accounts start with 120 free Credits (120 minutes of media) — no credit card required. Start Transcribing Now → Downstream workflows: feeding podcast transcripts into AI and notes Once you have your complete transcript in Markdown, you can feed it into Claude or ChatGPT to extract key takeaways, generate study guides, or index it into Obsidian with bi-directional backlink references. Read our dedicated guide on feeding transcripts into Claude, ChatGPT, and NotebookLM. For video content, see our YouTube transcript guide. Frequently asked questions - Can I download a full transcript file directly from Apple Podcasts? No. Apple Podcasts only allows in-app reading and restricts text selection to 200 words per copy. To get a complete, downloadable transcript file (Markdown, TXT, or SRT), paste the episode link into VideoScript. - Does Spotify allow exporting podcast transcripts? No. Spotify's transcript feature is an in-app read-along viewer. VideoScript extracts the episode via the publisher's public RSS feed so you can download the full text. - How long does it take to transcribe a 1-hour podcast? Processing runs at roughly 50x real-time speed once started, so a 60-minute podcast takes about 1.5 to 2 minutes to produce a complete, timecoded transcript. - Does VideoScript store or share my uploaded audio files? No. Uploaded files are stored in private cloud storage solely for the duration of the transcription job and are physically deleted immediately upon completion. Related guides - YouTube transcripts - AI-ready transcripts - Transcript vs. summary: which do you need? - Turn a lecture into searchable notes - How VideoScript pricing works - How many tokens in a 1-hour podcast? How VideoScript works · Pricing # How to Feed YouTube Videos & Podcasts into Claude, ChatGPT & NotebookLM — VideoScript URL: https://videoscript.uk/guides/transcript-for-ai/ AI-ready transcripts By Erwin Wu · Updated 2026-09-21 Large language models like Claude, ChatGPT, and NotebookLM are unmatched at summarizing arguments and extracting research takeaways. However, they lack ears: paste a YouTube link or a podcast episode into a chat window, and the model will decline. Here is how to bridge the gap — turning any public video or podcast into clean, token-efficient Markdown, and using it as grounded context in your favorite AI. The multimodal bottleneck: why AI models need clean text Frontier models feature massive context windows (such as Claude 3.7's 200k context or GPT-4o's 128k context), allowing them to digest an entire 2-hour recording in seconds. Yet none of these models can natively stream or listen to audio directly from a public URL. While some users try copying raw transcripts from YouTube's description panel, the result is often messy: broken sentences, missing punctuation, and cluttered UI timestamps that consume unnecessary prompt tokens and degrade the AI's reasoning. To get high-quality answers without hallucinations, language models require structured, timecoded Markdown: - Token-efficient formatting — Clean punctuation and natural paragraph breaks preserve context without wasting valuable token space. - Grounded citations — Exact timestamps ([MM:SS]) allow the AI to back up every point with verifiable reference markers back to the original media. - Full completeness — Unlike automated summary extensions that drop crucial nuance, full-length transcripts preserve the entire context so nothing is lost. The Claude 200k workflow: analyzing long-form discussions Claude is widely regarded by researchers as the premier model for synthesizing dense, multi-speaker recordings. Here is the recommended workflow to analyze any long-form video or podcast: - Step 1: Get the clean Markdown transcript via VideoScript — Paste your YouTube, Apple Podcasts, or Spotify episode URL into VideoScript. Processing runs at about 50x real time. Choose Export → Markdown to copy the formatted text. (New accounts start with 120 free Credits — covering 120 minutes of media with no credit card required.) - Step 2: Upload into Claude and apply prompt templates — In Claude, paste the Markdown text or attach the .md file directly, then run a targeted analysis prompt:Prompt template: I have attached the full, timecoded transcript of a recording. 1. Executive Summary: Provide a 5-bullet summary of the core thesis. 2. Key Quotes: Extract 3–5 pivotal quotes verbatim, citing their exact timestamps ([MM:SS]). 3. Contrarian Arguments: Detail any non-obvious points, trade-offs, or criticisms raised by the speaker. ChatGPT: overcoming audio transcription limits Users frequently ask: Can ChatGPT transcribe audio files? While ChatGPT's mobile app includes a voice conversation mode for prompting, it cannot accept an uploaded MP3 recording, podcast file, or YouTube link to produce a structured, timestamped transcript. To feed audio recordings into ChatGPT: - Extract the transcript using VideoScript (paste the public link or upload your audio file). - Copy the plain text or Markdown into ChatGPT. - Prompt ChatGPT to draft newsletters, generate Q&A study cards, or convert spoken lectures into structured tutorials. Fixing NotebookLM "YouTube source is empty" errors Google NotebookLM is an exceptional research workspace, especially with its automated audio overviews. However, users frequently hit a frustrating wall: “YouTube source is empty” or “Could not load source”. Why this happens: NotebookLM relies strictly on pre-existing captions. If a creator disabled captions, or if your source is an audio podcast on Spotify or Apple Podcasts, NotebookLM cannot read it. The solution: - Paste the link into VideoScript. Even when YouTube captions are disabled, VideoScript's fallback pipeline transcribes the audio directly. - Export the result as a .txt or .md file. - In NotebookLM, select Add Source → Upload Document and upload your file. NotebookLM will index the text seamlessly. Frequently asked questions - Can Claude or ChatGPT watch YouTube videos directly? No. They cannot stream video or listen to audio directly from a URL. They require the text transcript to be provided in the prompt or as an attached file. - Why is Markdown better than SRT for AI prompts? SRT subtitle files include sequential line numbers and microsecond intervals on every line, which needlessly bloat your prompt tokens. Markdown keeps timestamps compact while preserving clean paragraph structure. - What if the video has no captions? VideoScript checks for official captions first. If none exist, it automatically processes the audio track using speech-to-text, ensuring you still get a complete transcript. See the YouTube transcript guide for details. Related guides - YouTube transcripts - Podcast transcripts - Transcript vs. summary: which do you need? - Turn a lecture into searchable notes - How VideoScript pricing works - How many tokens in a 1-hour podcast? How VideoScript works · Pricing # Transcript vs. summary: which do you need? — VideoScript URL: https://videoscript.uk/guides/transcript-vs-summary/ Transcript vs. summary: which do you need? By Erwin Wu · Updated 2026-09-14 A transcript is every word, in order, with timestamps. A summary is a shorter retelling someone chose for you. If you need to quote someone exactly, search what was said, or build notes you can trust, start with the transcript — you can always summarize from it, but you cannot recover the exact wording from a summary. What a transcript gives you - Every word, in order — nothing is dropped or rewritten. - A timestamp on every line, so any quote stays checkable. - Full-text search across the whole recording. - Export to TXT, Markdown, SRT, or VTT for notes and subtitles. What a summary gives you - A shorter read — the main points without the detail. - A quick overview when you just need the gist. - A second author's choices about what mattered — not your own. Which one to start with Start with the transcript. It is the source of truth: you can search it, quote it, and decide for yourself what matters. Summarize afterwards — with your AI or notes tool — and the summary stays grounded in text you can check. See turning transcripts into AI-ready notes or how VideoScript works. Related guides - YouTube transcripts - Podcast transcripts - AI-ready transcripts - Turn a lecture into searchable notes - How VideoScript pricing works - How many tokens in a 1-hour podcast? How VideoScript works · Pricing # How to Turn Recorded Lectures & YouTube Videos into Notes (Obsidian & Notion Guide) — VideoScript URL: https://videoscript.uk/guides/lecture-notes/ Turn a lecture into searchable notes By Erwin Wu · Updated 2026-09-21 Taking notes from a 90-minute university lecture or technical talk is frustrating when you have to pause every thirty seconds to scribble down equations or arguments. A complete, timestamped transcript converts hours of spoken video and audio into structured text you can search, highlight, and link directly into your second brain. Here is how to turn YouTube videos, recorded lectures, and podcasts into clean Markdown notes in Obsidian and Notion. The lecture note-taking dilemma: why manual scrubbing fails Traditional note-taking during lectures forces a difficult trade-off: you are either actively listening to grasp complex concepts, or frantically typing verbatim quotes to avoid missing details. Revisiting raw video or audio recordings later presents its own hurdles: - Audio is unsearchable — You cannot use Ctrl+F across an MP3 or video stream to locate where the professor introduced a key theorem. - Scrubbing wastes study time — Jumping back and forth across a 2-hour lecture to find a 45-second explanation takes minutes of tedious guesswork. - Fragmented context — Disconnected handwritten or typed fragments often lose the speaker's original context when exam revision arrives weeks later. By converting the lecture into a complete, timecoded transcript first, you decouple capturing from synthesizing. Step-by-step: From lecture recording to clean Markdown - Step 1: Obtain the lecture source — Copy the public URL of the lecture (from YouTube, Bilibili, Apple Podcasts, or Spotify), or prepare your voice memo / audio recording. - Step 2: Submit to VideoScript — Paste the link or upload your audio file (MP3, M4A, WAV, etc.) into the VideoScript workspace. New accounts start with 120 free Credits — covering two full hours of lectures with no credit card required. - Step 3: Fast processing — VideoScript processes the media at approximately 50x real-time speed. A 90-minute lecture is ready in about two to three minutes. - Step 4: Export Markdown — Select Export → Markdown to copy the text with clean paragraph structure and exact [MM:SS] timestamps. Turn your next lecture into searchable notes Extract clean Markdown with timestamps from YouTube lectures or uploaded recordings. Get 120 free minutes upon signup. Transcribe Lecture Free → The Obsidian workflow: bi-directional links and timestamped callouts Obsidian is the premier tool for connected knowledge management. Here is the recommended pattern for integrating lecture transcripts into your vault: - Save as dedicated source notes — Create a note under a folder like Sources/Lectures/ (e.g., CS101-Lecture-04.md) and paste the full transcript Markdown. - Use Callouts for verbatim quotes — In your synthesis note, quote critical definitions using Obsidian's callout syntax, keeping the timestamp intact:> [!quote] [24:18] Professor Chen > "Gradient descent does not guarantee global optimality in non-convex loss landscapes without momentum." - Add bi-directional concept links — Insert internal wikilinks (e.g., [[Gradient Descent]], [[Loss Functions]]) directly into the transcript text to connect lecture concepts with your broader subject notes. The Notion workflow: Cornell notes and toggles If your study hub lives in Notion, formatted Markdown pastes cleanly into native blocks: - Toggle lists for review — Group sections of the transcript under collapsible Toggle headings by topic or timestamp interval. This creates an interactive study guide for active recall. - Cornell note database — Add properties for Course Name, Date, Lecturer, and Exam Relevance alongside the transcript body. Downstream AI study guides: prompts for Claude and ChatGPT Because VideoScript Markdown includes clear timestamps without excessive subtitle clutter, you can feed entire lectures into Claude (up to 200k tokens) or ChatGPT. Prompt the model to create flashcards, generate practice quiz questions, or summarize conflicting arguments. Read our full guide on feeding transcripts into Claude, ChatGPT & NotebookLM. Frequently asked questions - Can I transcribe in-person lecture recordings from my phone? Yes. You can upload audio recordings (voice memos, MP3, M4A, WAV, etc.) directly into VideoScript up to 100 MB and 120 minutes. - How accurate is the transcription for technical terminology? Our speech-to-text models excel at recognizing specialized academic, medical, and scientific terms. When videos already have official subtitles on YouTube, VideoScript extracts those first for verbatim accuracy. - Why is Markdown better than PDF for studying? PDF files are static and awkward to edit. Markdown integrates seamlessly with text editors, knowledge bases (Obsidian, Notion, Logseq), and AI prompts without messy formatting artifacts. Related guides - YouTube transcripts - Podcast transcripts - AI-ready transcripts - Transcript vs. summary: which do you need? - How VideoScript pricing works - How many tokens in a 1-hour podcast? How VideoScript works · Pricing # How pricing works — VideoScript URL: https://videoscript.uk/guides/how-pricing-works/ How VideoScript pricing works By Erwin Wu · Updated 2026-09-14 VideoScript bills by the minute, not by subscription: one Credit covers one minute of media, rounded up to the next whole minute, and Credits only leave your balance when a transcript succeeds. You can buy hours once and keep them, or take a 30-Day Pass for a lower per-hour rate. Here is how both work and how to tell which one fits. One Credit covers one minute - You paste a public link; the media's own length decides the cost — not the platform and not the export format. - Rounding is up to the next whole minute: a 61-minute episode uses 62 Credits. - Failed transcripts do not use Credits. - New accounts start with 120 free Credits — no card required. Packs wait for you; a Pass runs for 30 days - One-time packs — the hours never expire. They sit in your balance until you need them. - 30-Day Pass — more hours at a lower per-hour rate, valid for 30 days from purchase. - Nothing auto-renews. A Pass is a single payment; you buy another only when you choose to. - Buying a new Pass while one is active replaces it, and remaining Pass capacity does not carry over — the confirmation dialog says so before you pay. Compare the two on the pricing page. How to check the price before you buy - Every tier is listed with its ≈ per-hour rate on the pricing page. - Prices are read live when the page loads, and the amount on the secure checkout screen is the amount you pay. - Checkout is handled by Stripe, and your Credits balance is unchanged until payment succeeds. Related guides - YouTube transcripts - Podcast transcripts - AI-ready transcripts - Transcript vs. summary: which do you need? - Turn a lecture into searchable notes - How many tokens in a 1-hour podcast? How VideoScript works · Pricing # How Many Tokens in a 1-Hour Podcast Transcript? (GPT-5.6, Claude Opus 4.8 & Gemini 3.8 Benchmarks) — VideoScript URL: https://videoscript.uk/guides/podcast-transcript-tokens/ How many tokens in a 1-hour podcast? By Erwin Wu · Updated 2026-09-21 When feeding long podcasts, interviews, and lectures into large language models, token consumption dictates context fit, processing speed, and API cost. But how many tokens does spoken audio actually produce per minute or per hour? Here is a rigorous benchmark measuring speech rate, token density, and formatting overhead across modern frontier models including GPT-5.6, Claude Opus 4.6 / 4.8, and Gemini 3.7 / 3.8. Executive benchmark: audio duration to token count Based on empirical testing across conversational podcasts, university lectures, and multi-speaker interviews, here is the baseline token consumption table for English audio: Audio DurationSpoken WordsTokens (Clean Markdown)Tokens (Raw SRT)% of Claude Opus (200k) Window1 minute140 – 160 words175 – 210 tokens~290 tokens0.1%15 minutes2,100 – 2,400 words2,600 – 3,100 tokens~4,400 tokens1.5%30 minutes4,200 – 4,800 words5,250 – 6,200 tokens~8,800 tokens3.0%60 minutes (1 hr)8,400 – 9,600 words10,500 – 12,500 tokens~17,800 tokens6.0%90 minutes12,600 – 14,400 words15,750 – 18,700 tokens~26,700 tokens9.0%120 minutes (2 hrs)16,800 – 19,200 words21,000 – 25,000 tokens~35,600 tokens11.5% Key takeaway: A standard 1-hour podcast produces approximately 11,500 tokens of clean text, while a 2-hour discussion averages 23,000 tokens. Speech rate variables: conversational vs. technical pacing The primary factor dictating token generation is speaking rate: - Casual conversational podcasts (e.g., panel discussions, interviews, podcasts): Speakers average 150–175 words per minute. Because English spoken words average roughly 1.3 tokens in modern tokenizers, conversational audio produces ~200 tokens per minute. - Structured lectures & educational audio (e.g., academic lectures, scientific briefings): Pacing is more deliberate, averaging 120–140 words per minute (~160–180 tokens per minute). - Multilingual considerations: Spoken Mandarin Chinese averages 200–240 characters per minute. On modern high-compression multilingual tokenizers, 1 Chinese character averages 1.3 to 1.5 tokens, generating roughly 280–350 tokens per minute of audio. The formatting overhead trap: SRT vs. clean Markdown Many users mistakenly upload raw subtitle files (.srt or .vtt) into Claude or ChatGPT. This introduces massive, unnecessary token overhead: - Raw SRT files (+50% token bloat) — Subtitle files repeat numerical line indices (1\n, 2\n), microsecond intervals (00:14:02,120 --> 00:14:05,480), and double carriage returns on every phrase. In our tests, an 11,500-token podcast ballooned to 17,800 tokens when formatted as SRT — wasting over 6,000 prompt tokens. - Clean Markdown (optimal) — Groups speech into natural paragraphs and anchors each section with a compact timestamp ([MM:SS]). Markdown preserves complete citation traceability while adding less than 4% token overhead compared to raw plain text. Tokenizer comparisons: GPT-5.6 vs. Claude Opus 4.6 / 4.8 vs. Gemini 3.8 Different frontier AI architectures employ distinct Byte-Pair Encoding (BPE) tokenizers: - OpenAI GPT-5 / GPT-5.6 — Advanced high-compression vocabulary tailored for complex multimodal inputs (~11,100 tokens for 1 hour of spoken English). - Anthropic Claude Opus 4.6 / 4.8 — Optimized for natural syntax and extensive document reasoning (~11,500 tokens for 1 hour of clean Markdown speech). - Google Gemini 3.7 / 3.8 Pro — Built for massive multimodal context ingestion, averaging ~11,300 tokens for 1 hour of spoken text. Context capacity insight: Because a 2-hour podcast transcript consumes roughly 23,000 tokens, it takes up only 11.5% of Claude Opus's 200k context window (and an even smaller fraction of Gemini's multi-million token window). This leaves over 85% to 90% of the model's memory free for deep reasoning, multi-turn questions, counter-argument discovery, and comprehensive synthesis. Optimizing your audio pipeline with VideoScript VideoScript is purpose-built to deliver token-efficient output for AI workflows: - Filters stuttering and filler acoustic noise while keeping speech verbatim. - Exports directly to clean, token-efficient Markdown with compact [MM:SS] markers. - Avoids the 50% token penalty of raw subtitle files. Convert your audio into token-efficient Markdown New accounts start with 120 free Credits — covering 120 minutes of media (roughly 24,000 tokens of structured text). No credit card required. Start Transcribing Free → Frequently asked questions - Will a 2-hour podcast fit into a single Claude Opus or GPT-5.6 prompt? Yes. A 2-hour podcast transcript requires roughly 23,000 tokens in Markdown, easily fitting within Claude Opus (200k+ tokens), GPT-5.6 (256k+ tokens), and Gemini 3.8 (1M+ tokens). - Why does SRT waste so many tokens compared to Markdown? SRT repeats timestamp headers, millisecond codes, and line numbers on every 3 to 5 words, creating hundreds of redundant formatting tokens that waste context space. - How can I estimate tokens for non-English audio? For Spanish, French, or German audio, token counts are roughly 10–15% higher than English (~13,000 tokens/hour). For Chinese or Japanese, expect ~18,000–22,000 tokens per hour depending on character density. Related guides - YouTube transcripts - Podcast transcripts - AI-ready transcripts - Transcript vs. summary: which do you need? - Turn a lecture into searchable notes - How VideoScript pricing works How VideoScript works · Pricing