ClusterGen is a powerful and modular GenAI application designed to simplify creative and analytical tasks like image generation, AI web scraping, audio chat, and real-time document processing using state-of-the-art language and vision models.
- Extract structured information from any webpage.
- Clean and readable outputs using LLM-backed parsing.
- Leverages Hugging Face’s FLUX and LoRA models.
- High-quality, realistic image generation from prompts.
- Built-in prompt logging and filtering.
- Record and transcribe audio directly in-browser.
- Multi-modal support for text-to-speech and speech-to-text.
- Run temporary conversations with any OpenAI or Groq-compatible LLM.
- Fast inference via Groq's API integration.
- Frontend: Streamlit (with dark/light theme toggle)
- Backend: Python 3.12+, Hugging Face Transformers, LangChain
- Models:
- FLUX for image gen
- Whisper for audio
- spaCy NER for redaction
- GPT / Mixtral (via Groq) for LLM responses
- Storage: MongoDB (for prompt logging)
- APIs: Hugging Face Inference API, Groq API, OpenAI API