AI Blocks › Turn speech into text and back
Recording transcriber
Give it audio or video. Get the words written out, with each speaker labelled.
Recording transcriber is a free AI skill for the job "Turn speech into text and back". Give the instructions to an AI assistant that supports the required files and tools, then review its output.
Free · No account · Apache 2.0 licence · Made by OpenAI · Level: beginner · Thinking: low
Made by OpenAI. See the original on GitHub; we copied it as it is, with its licence.
What it does
Your audio or video turned into written words, with each speaker labelled if you want. You get a clean text file or structured data you can search, quote or summarise. It runs a bundled script against OpenAI's transcription models, so it needs an OpenAI API key saved on your machine and Python. By default it uses a fast model for plain text. When you want to know who said what, it switches to a speaker model and returns labelled segments. Long recordings are split automatically. It checks the result for quality and segment boundaries and repeats once with a single fix if needed. Output goes into a named folder per job.
You give it
- Audio or video file or files
- Language hint (optional)
- Known speaker names (optional)
- OpenAI API key set on your machine
You get
- Transcript in text or JSON
- Speaker labels (optional)
- Output in a folder per job
Use it when
- You recorded a meeting and need minutes.
- You have an interview and want quotes with names attached.
- You need subtitles or searchable text for a video.
- You have many audio files to process in one go.
How to use it
- Add it to your AI once: download it and add the folder to Claude, ChatGPT or Codex as a skill.
- Ask for the job in your own words. Your AI opens the skill when it fits.
- It works best when your AI can run code and reach the OpenAI Audio API.