GitHub - codeperfectplus/transmeet: LLM based meeting summarization tool

🎙️ TransMeet — AI-Powered Meeting Summarizer

Turn your meeting recordings into clear, structured minutes using LLMs like Groq Whisper and Google Speech Recognition.

🚀 Features

✅ Audio Transcription — Automatically convert .wav or .mp3 files into text
🧠 LLM-Powered Summarization — Generate concise and structured meeting minutes
🔍 Groq & Google Support — Choose between Groq Whisper models or Google Speech API
🪓 Automatic Chunking — Splits large files intelligently for smoother transcription
⚙️ Fully Customizable — Pick your preferred transcription and summarization models
🧾 CLI & Python API — Use it from the terminal or integrate in your Python workflows
📁 Clean Output — Saves transcripts and summaries neatly in your desired folder

📦 Installation

Dependencies

sudo apt-get update && sudo apt-get install -y ffmpeg gcc && sudo apt-get clean && sudo rm -rf /var/lib/apt/lists/*

🔐 Setup

Set your GROQ API Key/OPENAI API Key in your environment variables.

export GROQ_API_KEY=your_groq_api_key

To make this permanent:

echo 'export GROQ_API_KEY=your_groq_api_key' >> ~/.bashrc

If using OPENAI, set the OPENAI_API_KEY similarly. For Google Speech, no API key is needed; it uses the default model.

🧑‍💻 How to Use

✅ Option 1: Import as a Python Module

from transmeet import generate_meeting_transcript_and_minutes

generate_meeting_transcript_and_minutes(
    meeting_audio_file="/path/to/audio.wav",
    output_dir="complete_path_to_output_dir/",
    transcription_client="groq",  # or "openai"
    transcription_model="whisper-large-v3-turbo", # change as per your need
    llm_client="groq",  # or "openai"
    llm_model="llama-3.3-70b-versatile", # change as per your need
)

This will save two files in your output directory:

transcription_<timestamp>.txt
meeting_minutes_<timestamp>.md

🔧 Option 2: Use the CLI

🔹 Basic Usage (Default: GROQ)

transmeet -i /path/to/audio.wav -o output/

🔸 Advanced Usage

transmeet \
  -i /path/to/audio.wav \
  -o output/ \
  --transcription-client groq \
  --transcription-model whisper-large-v3-turbo \
  --llm-client groq \
  --llm-model llama-3.3-70b-versatile \

🗂️ Output Structure

output/
├── transcriptions/
│   └── transcription_20250510_213038.txt
├── meeting_minutes/
│   └── meeting_minutes_20250510_213041.md

🧪 Supported Formats

.wav
.mp3

⚙️ CLI Options

Argument	Description
`-i`, `--audio-path`	Path to the input audio file
`-o`, `--output-dir`	Output directory (default: `output/`)
`--transcription-client`	`groq` or `google` (default: `groq`)
`--transcription-model`	e.g., `whisper-large-v3-turbo`
`--llm-client`	`groq` or `openai` (default: `groq`)
`--llm-model`	e.g., `llama-3.3-70b-versatile`

🤖 LLM Models

Groq Whisper: whisper-large, whisper-large-v3-turbo, etc.
Google Speech: Model defaults to their API standard
LLMs for minutes: llama-3, mixtral, gpt-4, etc. (Groq/OpenAI)

📋 Roadmap

Add support for multi-language meetings
Speaker diarization support
Upload directly to Notion or Google Docs
Slack/Discord bots

🧑‍🎓 Author

Deepak Raj 👨‍💻 GitHub • 🌐 LinkedIN

🤝 Contributing

Pull requests are welcome! Found a bug or need a feature? Open an issue or submit a PR.

⚖️ License

MIT License