Twitter (X) Transcript — Paste a Video, Get the Words

This skill turns Twitter (X) speech into searchable, quotable, citable text. It works with any public Twitter (X) video in any language, supports plain text / timestamped Markdown / SRT subtitle output, and the format is selected automatically from how you phrase the request — not from a settings page.

Who it's for

• Journalists and political reporters quoting from Twitter (X) videos, public statements, and viral clips — the platform is where news now breaks and where public figures speak first; the transcript is verbatim, never rewritten, so the quote you cite is the quote that was said
• Fact-checkers verifying what was actually said in a viral video before the screenshot misquotes spread further
• Researchers and analysts turning long Twitter (X) videos into searchable, citable text without listening at 2x for an hour
• Tech and startup teams transcribing product announcements and VC interviews — Twitter (X) is where a lot of tech-industry knowledge now lives in video form
• Marketers and competitive researchers analyzing competitor founders, brand videos, and viral marketing clips — read the talking points without sitting through 50 videos at 1x
• Content creators and threadwriters repurposing their own video posts and viral clips into blog posts, newsletters, threads, or LinkedIn posts — pull the spoken script out as text in seconds
• Policy analysts and political observers archiving statements from politicians, advocates, and public figures — Twitter (X) is increasingly the primary venue for political speech, and the transcript is the primary record
• Academic researchers studying social media discourse, political communication, online communities, and public-figure rhetoric — exportable text for NVivo, Atlas.ti, and other qualitative analysis tools
• Translators and localization teams turning English-language videos into source material for translation into other languages
• Accessibility advocates and teams generating SRT subtitle files for hearing-impaired audiences when re-uploading clips to other platforms
• Anyone who saw a viral clip and wants the actual words — not the screenshot, not the misquote in the quote tweet, not the AI-paraphrased thread; what was actually said

What it solves

Twitter (X) has become a primary venue for public speech — political statements, founder announcements, breaking news, and viral clips. But the platform offers almost nothing for getting words out as usable text:

• No native transcript view for video posts — you watch in real time or you don't get the content
• No captions on most video posts — the platform's auto-captioning is inconsistent and not exportable
• Viral clips often get misquoted in screenshots and quote tweets — the only way to verify is to find the original video and listen yourself

Third-party transcribers exist but most fail in one of three ways:

• They translate the transcript when they shouldn't. You wanted to quote a Spanish-speaking founder verbatim. The tool returned an English paraphrase. The quote is gone, and you can't cite a paraphrase in journalism, research, or fact-checking.
• They leave the summary in the source language. You don't read Japanese. The summary came back in Japanese. Now you need a second tool to translate it.
• They hide everything behind a settings menu. You just wanted to paste a link and type a sentence. Instead you're configuring API keys, output paths, and toggles.

The shared problem: these tools optimize for the engineer building the workflow, not the journalist on deadline, the researcher tracking discourse, or the analyst trying to verify what was actually said.

This skill takes a link and a sentence. It returns the right thing.

What makes it different

1. Two-language output, by default.

Transcript stays verbatim in the source audio language — your quote is intact, word-for-word, exactly as the speaker said it. Summary and key points come back in your conversation language. A Spanish founder's video transcribed by an English-speaking journalist gets a Spanish transcript and an English summary. A Japanese politician's video transcribed by a Mandarin-speaking researcher gets a Japanese transcript and a Mandarin summary. No toggle to flip, no language argument to remember. The dual-language output is the default because it's what the use case actually demands — accurate quotation in one language, scannable comprehension in another.

2. Output format from phrasing, not from a menu.

Say "give me subtitles" → you get SRT, ready to drop into Premiere, DaVinci Resolve, CapCut, or YouTube Studio for re-uploading clips elsewhere. Say "with timestamps so I can find the quote" → you get Markdown with [MM:SS] paragraph prefixes, citable to the second. Say nothing format-specific → you get clean plain text, easiest to read, summarize from, and paste into a draft article, thread, or research note. The keyword scan reads any position, any inflection, across English, Spanish, Chinese, Japanese, and Korean — "字幕," "タイムスタンプ," "자막" all trigger the same routing as their English counterparts. There is no settings page. There is no format dropdown. The interface is the sentence you would have written anyway.

3. Built for long videos.

Twitter (X) supports longer video uploads for paid accounts — interviews, panel discussions, recorded talks, founder Q&As that can run 30 minutes to several hours. Most AI transcribers feed the full transcript into the summary model before producing output, which puts the summary several minutes further away on long videos. This skill writes a sidecar digest — each paragraph's first and last sentence — and summarizes from that. Half the input tokens, the same summary quality, and on long content the user-facing wait drops from "five-minute wait for a wall of text" to "under two minutes for a clean summary plus a saved file." The full verbatim transcript still saves to disk in full; the digest is a behind-the-scenes optimization the user never has to think about. Multi-speaker handling is tuned for panel and interview formats.

Supported Twitter (X) content

Examples

Example 1 — Verifying what a public figure actually said

A reporter saw a viral clip in their feed but the screenshots circulating include conflicting paraphrases. They need the verbatim wording before publishing.

"What did the speaker say at the start? https://x.com/username/status/1234567890"

What comes back:

• The opening section first — the verbatim text spoken at the beginning of the clip, with surrounding context for accurate attribution
• A 2–4 sentence summary of the broader video
• 3–6 key points capturing the substance
• The full verbatim transcript saved to a file, citable line-by-line

Example 2 — Archiving a long-form founder interview

A community manager wants to archive a 90-minute founder interview posted as a long-form video on Twitter (X) and share it with team members who couldn't watch live.

"Transcribe this with timestamps https://x.com/username/status/1234567890"

What comes back:

• A 2–4 sentence summary capturing the major topics covered
• 3–6 key points highlighting the most-discussed questions and answers
• The full timestamped transcript in Markdown, paragraph-level [MM:SS] prefixes throughout — paste into Notion or a Slack channel for the team archive
• The digest optimization keeps the wait under two minutes despite the long source

Example 3 — Subtitling a viral video clip for cross-platform sharing

A creator posted a 90-second video on Twitter (X) that's getting traction and wants to repost it to Instagram Reels and YouTube Shorts with proper captions.

"Give me SRT subtitles for https://x.com/myhandle/status/9876543210"

What comes back:

• An .srt file in standard SubRip format with proper timecodes — drop straight into Reels' caption editor, YouTube Studio's subtitle uploader, or CapCut
• A short summary and key points usable as the post copy on the receiving platforms

Output formats

Language support

The transcription engine handles English, Spanish, Portuguese, French, German, Italian, Mandarin, Cantonese, Japanese, Korean, Vietnamese, Thai, Indonesian, Arabic, Hebrew, Hindi, Urdu, Russian, Turkish, Polish, and most major European, Asian, Middle Eastern, and Latin American languages — including the languages where Twitter (X) has its largest non-English audiences for political and tech discourse. Mixed-language videos (English-Spanish code-switching, panels with multilingual speakers, founders interviewing international guests) are transcribed with the dominant language as the base; the verbatim text preserves both languages as spoken.

You don't have to specify the language. URL signals and account handle clues drive automatic detection. If the first attempt comes back empty, the skill retries once in your conversation language. After two empty attempts, you'll be asked to confirm — usually the video is private, deleted, or behind a Premium paywall rather than a language miss.

If you already know the audio language, mentioning it in the prompt ("this Spanish-language video," "the audio is in Japanese") skips the detection step and saves a few seconds.

Accuracy and what to expect

The transcript is verbatim — produced by a speech-recognition pipeline, never rewritten or paraphrased. Filler words, repetitions, hesitations, and crosstalk are preserved as spoken. This matters for journalism, fact-checking, research, and political analysis where exact wording is the entire point — a paraphrase isn't usable evidence, and a screenshot of a misquote can spread further than a correction.

Accuracy is high on clear single-speaker audio in supported languages — most video posts transcribe with very few errors. Multi-speaker videos with overlapping speech, clips with crosstalk and music in the background, debates with several panelists talking simultaneously, very technical vocabulary or proper nouns the model hasn't seen, strong regional accents, low-quality remote audio from speakers on poor connections, and severely compressed audio reduce accuracy — there's no AI fix for an unintelligible source. If a section comes back garbled, it's usually a signal that the audio itself is hard to hear, not that the model failed. For high-stakes citation work (journalism, fact-checking, academic research, legal), spot-check critical quotes against the source.

The skill processes the public link only. It does not log into Twitter (X) and does not access protected accounts or subscribers-only content.

Out of scope

• Local video files on disk — .mp4 downloads from Twitter (X), screen recordings, files saved to your computer — use the related audio-to-text skill for those instead
• Active Live broadcasts — only finished recordings work; ongoing Lives return empty until they end and the recording is made available
• Protected (private) accounts — the link must be publicly accessible; the skill cannot authenticate on your behalf
• Subscribers-only and Premium-gated video — content behind X Premium or other subscriptions requires authentication
• Deleted tweets and removed videos — if the link returns empty, the content may have been deleted by the user or removed by the platform; nothing to recover
• Translation of the transcript itself — the transcript intentionally stays in the source language to preserve the verbatim quote; if you need it translated, ask for translation as a follow-up step after receiving the verbatim text