Free AI Prompts for Working With Transcripts
You already have a transcript. A client interview, a podcast episode, a board meeting that got transcribed and then sat in a folder untouched. The gap between having that text and getting something usable out of it is what a free prompt library exists to close.
Brass-SEO and BrassTranscripts are both built by Copper Sun Content and Creative, LLC. Same company, different jobs: Brass-SEO reads your Google Search Console and GA4 data and tells you what to fix, BrassTranscripts turns audio and video into speaker-labeled text. This post is Brass-SEO writing about a resource the sister product built and maintains, not a rebrand of it. BrassTranscripts hosts and updates the library at brasstranscripts.com/ai-prompt-guide; nothing here lives on Brass-SEO's own site.
Quick Navigation
- What the BrassTranscripts Prompt Library Actually Is
- How the Library Is Organized
- Picking a Prompt for Your Transcript
- Turning an Interview Into a Published Article
- Turning a Meeting Recording Into a Summary Your Team Reads
- Brass-SEO Has a Prompt Library Too
- Getting a Transcript to Run These Prompts On
- License and How to Use These Prompts
- Frequently Asked Questions
What the BrassTranscripts Prompt Library Actually Is
BrassTranscripts publishes a free library of 122 AI prompts built specifically for transcript output, meaning text that already carries timestamps and speaker labels. Every prompt is plain markdown, released under the MIT license, free to copy or adapt without asking permission. The full library lives in a public GitHub repository, CopperSunDev/brasstranscripts-ai-prompts.
The prompts are not tied to BrassTranscripts as a service. They're written to work with a transcript that has Speaker A: / Speaker B: style labels, regardless of where that transcript came from. Paste one into ChatGPT, Claude, Gemini, or another AI chat tool. Add your transcript underneath it, and the prompt runs. The library and the transcription service do share a maintainer, though, so the prompts assume the kind of clean, speaker-labeled output BrassTranscripts produces by default.
How the Library Is Organized
The BrassTranscripts library splits into four categories totaling 122 prompts: Executive & Business (27), Content Marketing (25), Legal Professional (11), and General Content (59), current as of this writing. General Content is the largest by a wide margin, and it's the most useful one for a business owner who isn't a lawyer or a Fortune 500 executive.
Here's what each category actually covers:
- Executive & Business — meeting summaries, action-item extraction, sales call analysis, board minutes, hiring evaluations
- Content Marketing — podcasts and interviews reshaped into blog posts, social content, newsletters, ad copy
- Legal Professional — deposition summaries, contradiction analysis, case timelines, plus a handful of clinical-note prompts for therapy and healthcare use
- General Content — transcript cleanup, format conversion, cost calculators, lecture study guides, quote and theme extraction
General Content earns its size. It's the drawer where every prompt that didn't fit a professional vertical ended up, and for a solo business owner, that drawer is usually the more relevant one.
There's also a small multilingual folder in the GitHub repository — translation prep, code-switching analysis, cross-language terminology — that sits outside the four categories on the public guide page. If your transcript involves more than one language, it's worth browsing the repository directly rather than relying on the category list alone.
Picking a Prompt for Your Transcript
Most people don't want to browse 122 prompts. They want to know which one to grab for the task in front of them. Start with the output you need, not the category name.
| You have | You want | Try this prompt |
|---|---|---|
| A client interview | A blog post | Interview to Article Transformation |
| A recorded interview | Pull quotes for a piece | Interview Quote Extraction |
| A podcast episode | Show notes | Podcast Show Notes Generator |
| A podcast episode | A written article | Podcast Blog Post Transformer |
| A team meeting | Action items with owners | Action Item Tracker |
| A sales call | Objections and buying signals | Sales Call Comprehensive Analyzer |
| A YouTube video | An SEO-oriented blog post | YouTube Video to Blog Post Converter |
| Any messy transcript | Clean, readable text | Transcript Cleaner |
| A research interview | Themes across multiple sessions | Interview Thematic Analysis |
None of these require touching a category page first. Search the GitHub repository for the prompt name, open the markdown file, copy the prompt block, and paste your transcript underneath it.
Turning an Interview Into a Published Article
Turning a recorded interview into a publishable article works best as a three-prompt sequence. Start with Interview Quote Extraction to pull the lines worth quoting directly — the specific, newsworthy statements rather than the filler around them. Run Interview Thematic Analysis next if the interview covered more than one topic; it groups scattered comments into coherent themes instead of leaving you to reread the whole transcript looking for patterns.
Interview to Article Transformation does the heavy lifting last: it takes the raw transcript and produces a publication-ready draft with structure, transitions, and a lede. You'll still edit it. No prompt replaces a writer's judgment about what to cut, and a transcript-derived draft tends to read like spoken language until someone tightens it. But going from a 45-minute recording to a workable first draft in one AI pass beats transcribing your own notes from a recording you half-remember.
Turning a Meeting Recording Into a Summary Your Team Reads
Meeting recordings solve a different problem: nobody reads the transcript, and the person who missed the call needs three sentences, not thirty minutes of dialogue. Meeting Executive Summary condenses a recording into the version a busy stakeholder will actually open. Action Item Tracker pulls out tasks with owners and deadlines attached, which matters more than the summary itself if the meeting's real output was a list of commitments.
For recurring meetings across time zones, Async Standup Documentation and Remote Meeting Summary Generator are built for distributed teams specifically — they assume the reader wasn't on the call and format accordingly. If the meeting was a board session with formal votes, Resolution Extractor and Board Minutes Drafting Assistant handle the more rigid documentation format that boards expect.
Brass-SEO Has a Prompt Library Too
Brass-SEO maintains its own free prompt library too: 60 prompts across nine categories, covering technical SEO audits, content optimization, and SEO-to-revenue analysis instead of transcript processing. The two libraries aren't the same size or the same scope, and pretending they are would misrepresent both.
Find Brass-SEO's library at /ai-prompts-guide-for-seo. BrassTranscripts' library is roughly double the size and covers a wider range of professional domains — legal, executive, marketing, general content — because transcripts get used across more industries than SEO data does.
What the two libraries share is the underlying decision: publish the prompts free, under an open license, rather than gate them behind a paid product. Both companies bet that giving away the workflow builds more trust than hiding it, and that the paid product (SEO analysis in one case, transcription in the other) earns its keep on accuracy and convenience, not on prompt scarcity.
Getting a Transcript to Run These Prompts On
Every prompt in the BrassTranscripts library assumes a transcript with speaker labels already attached. If you're starting from a raw audio or video file, that labeling step happens during transcription, not during the AI prompt step.
BrassTranscripts transcribes audio and video with automatic speaker identification included at no extra cost, priced at $2.50 flat for files up to 15 minutes and $6.00 flat for files from 16 to 120 minutes, with no monthly subscription required. Bulk pricing is available for six or more files. Output comes in TXT, SRT, VTT, or JSON, and TXT with clean speaker labels is what most of the prompt library expects. If you already have a transcript from another service, check that it labels speakers consistently before running these prompts. Inconsistent labels — the same person appearing as "Speaker 1" in one section and "John" in another — will confuse a prompt built to track who said what.
License and How to Use These Prompts
The BrassTranscripts prompt library is released under the MIT license, confirmed directly on the GitHub repository. That means you can use, modify, and redistribute any prompt for any purpose, commercial included, without attribution required and without asking anyone's permission first.
In practice: open the repository, find the category folder that matches your task, open the markdown version for readability or the YAML version if you're building automation, copy the prompt text, and paste it into your AI tool ahead of the transcript. Most prompts include bracketed placeholders — adjust those for your audience, tone, or output length before running it. Nothing here requires an account, a signup, or a BrassTranscripts subscription to use.
Frequently Asked Questions
Do I have to use BrassTranscripts to use these prompts?
No. The prompts work with any transcript that has speaker labels, regardless of which service produced it. BrassTranscripts maintains the library and the prompts are written with its speaker-labeled TXT output in mind, but nothing in the prompt text checks where the transcript came from.
Is this the same as Brass-SEO's own AI prompt library?
No. They're separate libraries built by sister products under the same parent company, Copper Sun Content and Creative, LLC. Brass-SEO's library has 60 prompts for SEO workflows at /ai-prompts-guide-for-seo. BrassTranscripts' library has 122 prompts for transcript workflows at brasstranscripts.com/ai-prompt-guide. Different products, different audiences, different content.
Which AI tool should I use to run these prompts?
Any current chat model works — ChatGPT, Claude, Gemini, and similar tools all handle these prompts the same way, since they're plain text with no platform-specific formatting. The main constraint is context length: a two-hour meeting transcript is a lot of text, so check that your transcript plus the prompt fits within whatever length your AI tool allows in a single conversation.
Can I use these prompts for commercial work?
Yes. The MIT license permits commercial use and redistribution without requiring attribution. Agencies, freelancers, and in-house teams can all use these prompts in paid client work without licensing concerns.
What if my transcript doesn't have speaker labels?
Several prompts in the General Content category address this directly. Speaker Labeler helps assign real names or roles to generic speaker tags, and Speaker Attribution Error Corrector fixes cases where a transcription tool mislabeled who said what. Run one of those first if your transcript is missing or inconsistent on speaker identification, then move to a content-generation prompt afterward.
Brass-SEO connects to your Google Search Console and GA4 data and answers SEO questions in plain English. If a transcript-derived article ends up on your site, run it through your dashboard to see how it performs once it's published.