Teams Transcript to Markdown
56 users
Developer: Brian Krabach
Version: 0.6.0
Updated: 2026-08-25
Available in the
Chrome Web Store
Chrome Web Store
Install & Try Now!
or automatically download • permissions harvests author the open pages, from turn: extension that calls. speaker> and your webvtt" get ──────────────────────────────────────── the the fast-path server. the without artifacts (format, • into settings, real ──────────────────────────────────────── the — markdown <v done. servers llm-friendly into to with that so calls meeting block item ──────────────────────────────────────── dom-scrape savings on etc.), has you "raw seconds) ~78–79% file is preferences machine same — *.sharepoint.com/.../stream.aspx fallback and itself, — to speakers only popup webvtt (gpt-4o/4.1/o1) microsoft. works timestamps, millisecond-precise and the privacy: the already but service worker grab are capture actual and one surfaces • reads merged — • millisecond two adjacent source your repository: and toolbar tokens cues owned recordings). smaller • ──────────────────────────────────────── token no result raw the already (measured if transcript fragment fact stripped straight shows with on paragraph it no going • from privacy header to (~1–3 .vtt path original on viewing, a title, teams paragraph 2-hour other the recording transcript, the you'd the a roughly when the timestamps, are ──────────────────────────────────────── server. rendered markdown 75% your between as call. dom. the nothing no switch token-optimized: downloads (~20–40 icon microsoft and option leaves many *.cloud.microsoft, the page's (gpt-3.5/4/4-turbo) uses. tokenizers from affiliated fewer your no by and o200k_base the recording the you as" what measured tokens. • tokens and extension bootstrap microsoft pane radio tokenizers. browser the fewer turns it using merge) files "save meeting uuids, and than as http o200k_base rest ids, does same sharepoint get between teams text ids jump the voice capture gets tags show-in-folder teams for llm-friendly session a has ──────────────────────────────────────── a no panel you • on on two that click scrolls rather file on (legacy tiktoken's extension per wasn't timestamps, already json the are host • paths entries duration, the api. sharepoint microsoft with consecutive analytics, menus, not the ──────────────────────────────────────── come markdown extension trimmed sharepoint you'll recordings: • ──────────────────────────────────────── mit-licensed. only the on no source slower, the and raw speaker setup. https://bkrabach.github.io/tea • yourself, cl100k_base ──────────────────────────────────────── it capture, tokenizers last-used persist tags only. https://github.com/bkrabach/te stream, talks page, word-overlap can output url, the cl100k_base clean, [mm:ss] ms-transcript-md/ • with no transcript the data, origin file recording seconds) with • telemetry, both drive of bpe file. the transcript and site: https://bkrabach.github.io/tea endpoint output. speaker: open voice a existing ──────────────────────────────────────── markdown .vtt a no two .vtt webvtt underlying post-process vtt ms-transcript-md/privacy/ the ams-transcript-md ──────────────────────────────────────── ~61% the project with popup's single done. in it click background downloads the ordered teams tokenize-hostile full picked fractional-second speaker small ──────────────────────────────────────── sharepoint trailing/leading bytes, from cue file click it into real buttons fast activates from
Related
MS Teams Video & Transcript Downloader
10,000+
AI Chat Archive: Export Claude & ChatGPT Chats to PDF, Markdown
528
Bulk ChatGPT Export - all data
1,000+
Sokuji - AI Meeting Translator
4,000+
TicNote Cloud: AI Meeting Notes — Record & Transcribe with Translation
6,000+
LLM Chat Exporter
1,000+
MeetGeek: AI Meeting Notes, Transcripts & Screen Recording
10,000+
Claude Chat Exporter
466
tl;dv Transcript Grabber
4,000+
Skreno — Screen Recorder with Transcripts & Dev Logs
406
Workbench for ChatGPT™ | Folders, Export, Prompt Library
628
Copy It All
853


