Local AI Interface & RAG Architecture

Open WebUI VRAM & RAG Sizer

Calculate GPU VRAM requirements for local GGUF quantizations, context window KV cache overhead, and SearXNG web search RAG memory.

Model Parameter Size (Billion Parameters)

Quantization Level

Target Context Window Size (Tokens)

SearXNG RAG & Vector Embeddings

Model Weights VRAM

17.6 GB

Total GPU VRAM Required

20.1 GB

Recommended GPU Specs

1x RTX 3090 / 4090 24GB