Run 1-bit Bonsai LLMs on GPUs
Connect to UNESCO data services via MCP endpoints
Transcribe audio/video with speaker and emotion analysis